ArticlePDF Available

Water flow controls distribution and feeding behavior of two co-occurring coral reef fishes: II. Laboratory experiments


Abstract and Figures

The chaenopsid blenny Acanthemblemaria spinosa occupies topographically high locations on coral reefs where flow speeds and turbulence are frequently greater than those experienced by its congener, A. aspera, which occupies locations close to the reef surface. To investigate the adaptive mechanisms resulting in this microhabitat differentiation, the foraging effort and success of these fishes were determined in laboratory flumes that produced flow conditions approximating those experienced in the field. Individual fish were subjected to unidirectional (smooth and turbulent) and oscillatory flows while they fed on calanoid copepods, Acartia tonsa, whose vulnerability to predation varies with water flow. In unidirectional flow both blenny species had their greatest foraging success at intermediate flow speeds (ca. 10cms−1) and under turbulent flow. Under all conditions, Acanthemblemaria spinosa exhibited greater foraging effort and attacked at greater distances, greater mean water speeds, and in oscillatory flow, over a greater proportion of the wave cycle than did A. aspera. A. spinosa also exhibited greater foraging success under turbulent flow conditions. These differences in feeding patterns allow A. spinosa, with its higher metabolic rate, to occupy the more energetic higher locations in corals where planktonic food is more abundant. A. aspera occupies the poorer quality habitat in terms of planktonic food availability but its lower metabolic rate allows it to thrive there. Consequently, these species divide the resource in short supply, i.e., shelter holes, based on their differing abilities to capture prey in energetic water conditions in conjunction with their differing food energy requirements.
Content may be subject to copyright.
Water flow controls distribution and feeding behavior of two
co-occurring coral reef fishes: II. Laboratory experiments
R. D. Clarke ÆC. M. Finelli ÆE. J. Buskey
Received: 1 August 2008 / Accepted: 7 February 2009 / Published online: 4 March 2009
ÓSpringer-Verlag 2009
Abstract The chaenopsid blenny Acanthemblemaria
spinosa occupies topographically high locations on coral
reefs where flow speeds and turbulence are frequently
greater than those experienced by its congener, A. aspera,
which occupies locations close to the reef surface. To
investigate the adaptive mechanisms resulting in this
microhabitat differentiation, the foraging effort and success
of these fishes were determined in laboratory flumes that
produced flow conditions approximating those experienced
in the field. Individual fish were subjected to unidirectional
(smooth and turbulent) and oscillatory flows while they fed
on calanoid copepods, Acartia tonsa, whose vulnerability
to predation varies with water flow. In unidirectional flow
both blenny species had their greatest foraging success at
intermediate flow speeds (ca. 10 cm s
) and under tur-
bulent flow. Under all conditions, Acanthemblemaria
spinosa exhibited greater foraging effort and attacked at
greater distances, greater mean water speeds, and in
oscillatory flow, over a greater proportion of the wave
cycle than did A. aspera.A. spinosa also exhibited greater
foraging success under turbulent flow conditions. These
differences in feeding patterns allow A. spinosa, with its
higher metabolic rate, to occupy the more energetic higher
locations in corals where planktonic food is more abundant.
A. aspera occupies the poorer quality habitat in terms of
planktonic food availability but its lower metabolic rate
allows it to thrive there. Consequently, these species divide
the resource in short supply, i.e., shelter holes, based on
their differing abilities to capture prey in energetic water
conditions in conjunction with their differing food energy
Keywords Water flow Plankton capture Planktivore
Turbulence Copepod
Wave energy varies considerably across coral reefs (Gourlay
and Colleter 2005; Kench 1998; Madin et al. 2006; Young
1989) and the varying water motion plays an important role
in the spatial relations of fishes ranging from large-scale
habitat distribution across major reef zones (Bellwood and
Wainwright 2001; Fulton and Bellwood 2005) to small-scale
distribution among coral heads (Johansen et al. 2008). The
scale of the effect is largely determined by the size of the
fishes (Depczynski and Bellwood 2005), their swimming
abilities (Fulton et al. 2005), and the ways they use the reef
structure (Johansen et al. 2008). Whereas large active fishes,
such as parrot fishes, experience some reduction of the pre-
vailing surge by remaining within the benthic boundary layer
(1–2 m above the bottom) small cryptic fishes can benefit
from the greater water flow reduction due to the momentum
boundary layer a few centimeter above the bottom (Johansen
et al. 2008).
The effects of water flow on fishes extend beyond
potential dislocations and the energetic costs of swimming.
Communicated by Biology Editor Dr. Mark McCormick
R. D. Clarke (&)
Department of Biology, Sarah Lawrence College, Bronxville,
NY 10708, USA
C. M. Finelli
Department of Biology and Marine Biology, University of North
Carolina Wilmington, Wilmington, NC 28403, USA
E. J. Buskey
University of Texas Marine Science Institute, Port Aransas,
TX 78373, USA
Coral Reefs (2009) 28:475–488
DOI 10.1007/s00338-009-0479-7
For planktivores, it plays a significant role in their ability to
capture prey. Fast flow may make rapidly moving plankton
difficult for fishes to capture (Kiflawi and Genin 1997), but it
may also interfere with the ability of plankton such as
copepods to detect predators and escape (Robinson et al.
2007). For example, Clarke et al. (2005) showed that
switching from still to turbulent water conditions resulted in
a reduced capture success of non-evasive prey (Artemia
nauplii) and increased capture success of evasive prey (cal-
anoid copepods). The magnitude of these opposite effects
will vary with flow velocity and degree of turbulence.
Planktivorous fishes can be an important pathway for the
movement of carbon and nutrients from the water column
to the benthic community (Liberman et al. 1995; Hamner
et al. 1988; Porat and Chadwick-Furman 2005; Holbrook
et al. 2008). Benthic planktivores are an effective compo-
nent of this pathway because they deposit their feces and
dissolved nitrogenous wastes in close proximity to deposit
feeders and the absorptive surfaces of algae and corals. Not
only do these nutrients have shorter distances to diffuse,
but they also are released into the boundary layers near
surfaces and are less likely to be quickly dispersed back
into the water column. Chaenopsid blennies are incon-
spicuous but frequently abundant members of the benthic
community. Members of the genus Acanthemblemaria are
among the most closely tied to the substrate, to the point of
being called hemisessile (Kotrschal and Lindquist 1986).
They reach densities of 5 fish m
over large reef areas
(Clarke 1996) and can achieve local densities as high as
60 fish m
(Lindquist 1985). Thus, these fishes may be
overlooked contributors to nutrient input on coral reefs. In
addition, their small size, limited mobility and adaptability
make them excellent laboratory organisms that could shed
light on the dynamics of other planktivorous fishes that are
more difficult to study.
Acanthemblemaria spinosa and A. aspera live in holes
in corals from which they make rapid darts to capture small
crustaceans both on surfaces and in the water column. They
are the same size (generally 21–25 mm SL) and A. spinosa
has a standard metabolic rate 55% higher than that of
A. aspera (Clarke 1996,1999). A. spinosa generally lives
[0.5 m above the reef surface and A. aspera generally
lives \0.2 m above the reef surface (Clarke 1996). Con-
sequently, A. spinosa generally experiences greater flow
speeds and turbulence levels than A. aspera (Finelli et al.
2009); plankton density is also greater in the A. spinosa
microhabitat which is further above the reef surface
(Clarke 1992; Yahel et al. 2005). These differences cor-
relate with a larger proportion of A. spinosa feeding strikes
being in the water column (Finelli et al. 2009) and a greater
proportion of calanoid copepods in its gut (Clarke 1999).
Indeed, A. spinosa feeds primarily on planktonic prey and
A. aspera feeds primarily on benthic prey (Clarke 1999).
The hypothetical effect of water movement on the
feeding behavior of these fishes is depicted in Fig. 1. The
energy intake for planktonic feeding (foraging
effort 9foraging success 9mean caloric value of prey) is
represented by dome-shaped curves S (A. spinosa) and A
(A. aspera) (MacKenzie et al. 1994). The energy intake for
benthic feeding (E) is represented by a straight line with
negative slope because the momentum boundary layer
thins as water motion increases and benthic prey vulnera-
bility is unaffected by water motion, but fish swimming
ability is. The origin of the benthic feeding line is lower
than the plankton curves because the energetic value of
harpacticoids is likely to be lower than calanoids and
cyclopoids (Clarke 1999; White and Warner 2007) and
because harpacticoids are not delivered at the same high
rate as are planktonic forms. The same line is used for
benthic feeding by both fish species because there is no
information on how these rates may differ between them.
Dashed lines S and A represent the minimum energy intake
needed for maintenance, with S being higher than A
because A. spinosa has a higher standard metabolic rate
than does A. aspera (Clarke 1996,1999).
This model suggests several important implications of
water flow for the performance and distribution of these
species. First, point B represents the magnitude of water
motion at which the planktonic energy intake of A. aspera
equals that of benthic energy intake and is the point at
which this fish switches from planktonic foraging to
Water motion
Energy intake
Fig. 1 Model of energy intake versus water motion (speed and
turbulence) for Acanthemblemaria aspera and A. spinosa. Solid lines
S and A represent energy intake from planktonic feeding for
A. spinosa and A. aspera, respectively, and dashed lines S and A
represent their respective minimum energy requirements. Solid line E
represents energy intake from benthic feeding for both species
476 Coral Reefs (2009) 28:475–488
benthic foraging (Finelli et al. 2009). There is no similar
point for A. spinosa, for which planktonic energy intake is
always higher than benthic energy intake. Additionally, the
model suggests that as water motion increases, there is a
point for each species (point C for A. aspera and point D
for A. spinosa) beyond which they are not able to meet
maintenance needs through benthic or planktonic feeding.
This point occurs at lower flows for A. aspera than for
A. spinosa because of the greater ability of A. spinosa to
forage under more energetic flow conditions. The locations
of these curves will vary with different prey densities
within (Yahel et al. 2005) and between reefs (Clarke
1992). However, the model suggests that the spatial
distribution of these fishes is intimately tied to water
motion via effects on feeding performance and energy
The purpose of this study was to determine the association
between water flow and feeding performance of these two
fishes in laboratory flumes that simulated the periodicity,
velocity, and turbulence of water flow measured in the
feeding zones of these two blennies by Finelli et al. (2009).
Based on the conceptual model (Fig. 1), it was expected that
A. spinosa would demonstrate greater feeding success under
rapid water motion than A. aspera. Additionally, A. spinosa
was expected to be able to forage over a greater proportion of
the oscillating velocities of the wave cycle. Consequently,
A. spinosa would have access to more planktonic food than
A. aspera as depicted in Fig. 1. The significance of this is that
these fishes, by occupying different microhabitats, would
partition shelter holes, the resource in short supply, and thus
avoid competitive exclusion.
Materials and methods
Predator and prey species
During each of 3 years, 16 males and 16 females of
Acanthemblemaria spinosa and A. aspera were captured on
Glover’s Reef, Belize, and transported to the laboratory in
Port Aransas, Texas. The fishes were maintained in aquaria
as described in Clarke et al. (2005). Each fish resided in an
artificial shelter consisting of a 3 cm cube of Sculpey
polymer clay with a cylindrical cavity of 4.2 mm in
diameter and 25 mm deep located at the center of one face.
The fishes were fed Artemia sp. nauplii several times daily.
Zooplankton were captured with a 153-lm mesh
plankton net deployed either during mid-ebb tide from the
University of Texas Marine Science Institute pier in the
Aransas Ship Channel on the coast of Texas (27
030W) or hand towed in a canal on the west side of
Mustang Island, Texas (27
480N, 97
050W). Samples were
diluted in whole seawater and maintained in a plastic
bucket of mixed plankton with aeration and used within
18 h of capture. Individuals of the calanoid copepod
Acartia tonsa were sorted from the mixed plankton using
Pasteur pipettes and accumulated in beakers before being
placed in the flumes.
Since Acartia tonsa is not a natural prey of the blennies,
the effects of turbulence on the escape behavior of the
dominant calanoid copepod found in the natural environ-
ment, Acartia spinata, were also tested to ensure that
A. tonsa was a suitable substitute. A. spinata was captured
at 8 to 10 m depth on the forereef at Glover’s Reef
560N, 87
480W) with a diver-towed 0.25-m diameter
plankton net (153-lm mesh). Individuals were picked out
of the mixed plankton with a Pasteur pipette and used in the
turbulence tank experiments within 6 h of capture.
Effect of turbulence on prey behavior
In order to test the effects of turbulence on the escape
behavior of Acartia spinata, the prey species, the same
apparatus and procedures as those described in detail in
Gilbert and Buskey (2005) were employed. A small, clear
acrylic plastic observation chamber (20.5 910 915 cm)
housed a turbulence-generating apparatus consisting of an
oscillating, vertically oriented, plastic grid with 1.5 cm
openings, powered by an electric motor. The flow field near
a small siphon (2 mm inside diameter) was used to simu-
late a predator and thus stimulate escape behavior in the
copepods. The siphon emptied into a 1,000-mL beaker
containing 500 mL of water at 34 mL min
, and was
pumped back into the turbulence chamber behind the grid
at the same rate, using a peristaltic pump. The grid oscil-
lated between 7.5 and 9.5 cm from the siphon tip every
1.25 sec (0.8 Hz). A fiber optic light guide was used to
produce a narrow vertical shaft of light over the siphon tip
and the area in front of the siphon. Only copepods within
this light shaft were illuminated and recorded by a mono-
chrome video camera in a dimly lit room, limiting the
depth of field of the camera. This light shaft attracted
copepods to the vicinity of the siphon and away from direct
contact with the oscillating grid.
For both the turbulent and still-water treatments, 50 to
100 copepods were allowed to adjust to experimental
conditions in still water for five minutes after transfer to the
chamber. For the turbulent treatment, copepods were
allowed an additional five minutes in turbulence before
video recording began. The copepods were then recorded
for 10 min. Escape reaction distances and positions were
measured using Metavue image analysis software (Golden
Software). Escape reactions were defined as sudden chan-
ges in orientation followed by rapid swimming behavior
that normally caused the copepods to leave the field of
Coral Reefs (2009) 28:475–488 477
Effect of unidirectional flow speed on feeding
In order to test the effect of unidirectional flow speed on
feeding efficiency, a flume was used. It consisted of two
parallel straight sections (50 cm long) joined at the ends by
semicircles; the cross section of the channel was 10 cm
wide 915 cm high (Fig. 2). The water depth was main-
tained at 10 cm, resulting in a water volume in the flume of
17 L. A vertical paddle wheel in one straight section pro-
vided flow with minimal agitation, thus avoiding damage to
the copepods. The paddle wheel was driven by an electric
motor regulated by a Dart Micro-Drive II controller. The
other straight section comprised the working section of the
flume, at the upstream end of which was positioned a flow
straightener consisting of a 5-cm-thick baffle of 1 cm 9
1 cm plastic grid. A cube-shaped depression (3 cm on a
side) was built into the back wall and centered 5 cm above
the flume bottom and 35 cm from the upstream end of the
working section; this served to hold the blenny shelter with
the hole-bearing face flush with the flume wall. A stable
optical surface above the blenny shelter was provided by a
9cm915 cm piece of acrylic held in place at the water
Blennies were not fed on the days they were tested.
Tests were started by adding 85 Acartia tonsa to the flume
and allowing it to run for several minutes to disperse the
plankton and acclimate them to the flowing water. The
flume was then turned off, a blenny in its shelter was added
to the cube-shaped depression in the back wall of the flume
working section, the optical surface was positioned, and the
flume was started up. The experiment continued until 25
attacks were counted or 0.5 h had passed. Subsequent runs
were preceded by the addition of 25 copepods to make up
for those consumed. Because each approach did not result
in a capture, this was not an exact maintenance of copepod
density, but it kept the density at 5 to 6 copepods L
Although some copepods remained in the flume for up to
three runs, we are confident that they remained healthy and
responsive because previous tests showed no ill effects
under much stronger turbulence than used here (Clarke
et al. 2005).
Two sets of experiments were run in the unidirectional
flume. In the first experiment (Flow Speed Experiment,
Year 1), the effects of current speed alone were tested. Six
trials, each consisting of six randomized runs (two species
at three speeds), were performed. Each fish was used once.
Velocity measurements showed that turbulence increased
with flow speed. In order to separate the effects of turbu-
lence and flow speed, a second experiment was performed
(Turbulence Experiment, Year 2), where additional turbu-
lence at each flow speed was created by placing a dead
piece of the branched coral Porites porites in the center of
the working section 11.5 cm upstream of the blenny shel-
ter. A minimal amount of sculpting clay was used to hold
the coral in place. The coral and its base comprised 40% of
the cross sectional area of the working part of the flume.
The Turbulence Experiment consisted of eight trials each
consisting of 8 randomized runs (two species, two speeds,
two flow regimes). Twenty-five fish were used twice but
never in the same flow conditions.
Effect of oscillating flow on feeding efficiency
Because blennies in nature are continuously exposed to
wave-generated surge (Finelli et al. 2009), a flume was
constructed to reproduce that motion to observe predatory
behavior in oscillating flow. This flume consisted of a
working section with 10 cm 910 cm cross section sealed
with a transparent acrylic cover (Fig. 3). Each end was
connected via a 90°elbow to a vertical PVC pipe (15 cm
diameter). Flow was driven by a piston attached to an
electric motor regulated by a Dart Micro-Drive II con-
troller. The rotational motion of the motor was translated to
linear motion in the piston using a scotch yoke mechanism.
The points of slack water flow were indicated on the video
recordings by briefly switching on a light emitting diode
(LED) at the highest and lowest point of piston travel. By
controlling the rotation rate of the fly-wheel and the loca-
tion of the pin connecting the fly-wheel to the piston rod,
the period and amplitude (thus flow speed) of the oscil-
lating water in the working section could be regulated. The
water volume of the flume was 21.5 L. The period
remained constant at 5 s, the predominant period measured
in the field (Finelli et al. 2009).
Tests were started by adding a blenny in its shelter and
108 Acartia tonsa to the flume (=5 copepods L
) and then
Fig. 2 The unidirectional flume
478 Coral Reefs (2009) 28:475–488
sealing the observation section. Blennies were not fed on
the days they were tested. Each run consisted of three trials
at different flow speeds performed on the same fish; 25
copepods were added following a trial consisting of [15
approaches. In Year 3, six runs for each species were
carried out, with each possible sequence of three flow
speeds occurring once; thus, feeding data were recorded for
every flow speed twice at each of the first, second, and third
positions in the sequence, equalizing any effect of satiation
on feeding behavior among all flow speeds. Since
A. aspera exhibited low foraging effort at the middle and
highest speeds, additional trials were run to increase the
total number of approaches to improve statistical power
(seven at medium and 14 at high speeds).
Flow measurement
Current speeds and turbulent kinetic energy in both flumes
were measured using an Acoustic Doppler Velocimeter
(Sontek Micro ADV). The ADV measured all three compo-
nents of water velocity (u=downstream, v=cross-stream,
w=vertical) in a small volume (ca. 0.09 cm
) located 5 cm
from the probe. The ADV probe was positioned so that it
measured flow in the midline of the flume channel directly in
front of the blenny shelter. Measurements were made at
20 Hz. Average current speed in the unidirectional flume
was estimated as the mean of the vector sum of the three
velocity components (i.e., speed ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Because there was very little energy in the vand wcompo-
nents of velocity, average flows in the oscillating flume were
approximated as U
, the root mean square (standard
deviation) of the velocity time series in the downstream (x)
direction. For comparison, the mean flow speeds in the
oscillating flume as measured by the vector sum were *10%
lower than U
Measurements of turbulent water flow consist of mean
and variance components such that:
where uis the measured velocity,
uis the time-averaged
mean, and u0is the instantaneous deviation from the mean.
In order to estimate turbulence parameters, a modified
Fig. 3 The oscillating flume
Coral Reefs (2009) 28:475–488 479
Reynolds decomposition was used to separate the mean
and variance components of the three velocity signals (u, v,
w). For each velocity component, the mean velocity was
subtracted from each instantaneous measurement in the
time series (Toomas 2001). The turbulent kinetic energy
(TKE) was then calculated as
TKE ¼0:5u0u0þv0v0þw0w0
In the unidirectional flume, the mean velocity was
estimated as the arithmetic mean from the entire 5-min
record. In the oscillating flume, an appropriate mean was
estimated by subjecting the raw time series to a zero-phase
displacement five-point running average (filtfilt function,
Matlab v.14, Mathworks).
Image recording and analysis
For both the flumes, each run was recorded on VHS vid-
eotape (30 frames s
) with input from two cameras
combined into a split screen. One camera was facing
downward and captured a field of ca. 65 945 mm (x-y
plane), whereas the other camera, oriented horizontally,
faced the shelter and captured a field of ca. 70 960 mm
(x-z plane). The fields were lit with general room fluores-
cent lighting at 3.6 lmol photons m
for the
unidirectional flume and 2.4 lmol photons m
for the
oscillating flume. Feeding trials at six light intensities
ranging from 0.0002 to 21.6 lmol photons m
cated that the blennies attacked some nearby prey at
0.02 lmol photons m
and showed a full response at
0. 2 lmol photons m
. The fields were backlit with
infrared LEDs arranged to provide dark field illumination
so the video cameras could record the copepods. Fishes are
generally insensitive to infrared light (Lithgoe and Par-
tridge 1989).
Videotapes were viewed on a videocassette recorder
with variable speed and single frame advance. The first 20
approaches in each run were captured as ca. 3 s digital
clips and stored on a computer. These clips were opened
with Tracker Version 1.3 (Open Source Physics Java Video
Analysis) and the result of each approach (abandon, attack,
capture) was determined. For each attack, maximum
extension of the blennies was measured in three dimen-
sions using the center of the shelter opening as the origin
and the center of the upper jaw as the fish location.
Fish initially reacted to passing prey by directing their
eyes toward the copepod, swinging their heads to follow
the copepod, moving toward the copepod (approach), and
attacking the copepod. The reaction could end at any point
in this sequence. Attacks frequently led to high speed
escape jumps by the copepod; so capture was determined
by the disappearance of the copepod from the video frame
without its sudden reappearance elsewhere in the following
frame. Virtually all unsuccessful attacks resulted from such
escape jumps. An approach was considered as an intention
to feed and the failure to attack generally was prompted by
an escape jump or simply the rapid movement of the prey
as it was carried by the water; so foraging success was
measured as the proportion of approaches that led to cap-
tures. The number of approaches per minute was used as a
measure of foraging effort.
Statistical analysis
The Flow Speed Experiment and the oscillating flume data
were analyzed with two-way ANOVA for independent
samples with species (two levels) and speed (three levels)
being fixed factors. Single species were tested in the
oscillating flume experiment using one-way ANOVA for
independent samples with speed being a fixed factor (three
levels). The Turbulence Experiment was analyzed with a
three-way ANOVA for independent samples with species
(two levels), speed (two levels), and regime (smooth-tur-
bulent, two levels) being fixed factors. All ANOVAs
resulting from the unidirectional flume studies contained
equal sample sizes in all blocks and were normally dis-
tributed (Lilliefors Test for Normality). The oscillating
flume experiment resulted in larger sample sizes in two of
the six blocks but variances were homogeneous and normal
in three cases, and two potential ANOVAs were not per-
formed because of heterogeneous variances (Bartlett’s
Effect of turbulence on prey behavior
When compared to its performance in still water, Acartia
spinata in turbulent conditions reacted to the siphon at a
smaller distance and was captured at a greater rate
(Table 1). The values were very similar to those for Acartia
tonsa in the same apparatus; reaction distances differed by
\1% and 4% and the percent captured differed by 2% and
1% in still and turbulent water, respectively (Table 1).
A. spinata is a common congeneric copepod on Glover’s
Reef and at Carrie Bow Caye, where it is a significant part
of the diet of A. spinosa (Clarke 1999); thus A. tonsa is a
suitable stand-in for a portion of natural blenny prey.
Effect of unidirectional flow speed on feeding
A. spinosa exhibited a higher foraging effort than A. aspera
at all flow speeds (Figs. 4,5). The flow conditions achieved
in the Flow Speed Experiment are given in Table 2and
480 Coral Reefs (2009) 28:475–488
those in the Turbulence Experiment are given in Table 3.
Average flows in the flumes fall in the same range as
measured in fish feeding volumes in the field (5 cm s
15 cm s
for the modal wave heights at the different sites,
Finelli et al. 2009). For both the species, foraging effort
was lowest at the highest water speed (26.7 cm s
), but
for A. spinosa, it increased with mean water speed between
3.4 cm s
and 9.6 cm s
, whereas for A. aspera,it
decreased between these speeds in both the Flow Speed
Experiment (Fig. 4a; F
=8.42, P=0.007 for species;
=9.72, P=0.0006 for water speed; F
P=0.02 for species–water speed interaction; two-way
ANOVA) and the Turbulence Experiment (Fig. 5a;
=8.68, P=0.006 for species; F
P=0.26 for water speed; F
=5.07, P=0.03 for
species–water speed interaction; two-way ANOVA). A.
spinosa attacked at greater distances than A. aspera at all
speeds and each species attacked at the shortest distance at
the highest water speed in the Flow Speed Experiment
(Fig. 4b; F
=13.9, P=0.0008 for species;
=48.9, P\0.0001 for water speed; F
P=0.20 for species–water speed interaction; two-way
ANOVA) and the Turbulence Experiment (Fig. 5b;
=19.85, P\0.0001 for species; F
P=0.74 for water speed; F
=9.43, P=0.005 for
species–water speed interaction; two-way ANOVA). For-
aging success did not differ between species but was
significantly reduced at the highest water speed (Fig. 4c;
=0, P=1.0 for species; F
=4.22, P=0.025
for water speed; F
=0.42, P=0.66 for species–water
speed interaction; two-way ANOVA). Under smooth flow
conditions in the Turbulence Experiment, both species had
greater foraging success at 9.6 cm s
than at 3.4 cm s
(Fig. 5c; F
=0.42, P=0.52 for species; F
P=0.012 for water speed; F
=0.09, P=0.77 for
species–water speed interaction; two-way ANOVA).
The turbulence levels in the unidirectional flume were
increased by a factor of 8 to 10 with the placement of the
coral turbulence generator upstream of the blenny feeding
zones (Table 3). The achieved values were at the low end
of the range observed in the field (1 to 10 cm
at modal
wave heights at different sites, Finelli et al. 2009).
Turbulence had no effect on foraging effort (Fig. 5a;
=0.52, P=0.47 for flow regime; F
P=0.64 for flow regime–species interaction; F
P=0.30 for flow regime–water speed interaction; three-
way ANOVA) or attack distances (Fig. 5b; F
P=0.51 for flow regime; F
=0.33, P=0.57 for flow
regime–species interaction; F
=1.43, P=0.24 for
flow regime–water speed interaction; three-way ANOVA)
but significantly increased foraging success for both the
species (Fig. 5c; F
=12.0, P=0.001 for flow regime;
=3.5, P=0.07 for flow regime–species interaction;
=1.5, P=0.23 for flow regime–water speed inter-
action; three-way ANOVA). Indeed, foraging success was
enhanced by both greater turbulence and greater flow speed
(Fig. 5c; F
=24.5, P\0.0001 for flow speed;
=0.00, P=1.0 for flow speed–species interaction;
=1.5, P=0.23 for flow regime–water speed inter-
action; three-way ANOVA). Foraging success under
smooth flow was the same for both the species (Fig. 5c;
Table 1 Acartia spinata, distance from siphon tip at which escape reactions were initiated, the distance covered by the escape jump and the
percent of copepods captured by the siphon, under still water or turbulent conditions
Acartia spinata Acartia tonsa
Still Turbulent P Still Turbulent P
Reaction distance (mm) 5.48 (0.51) 3.69 (0.23) 0.029 5.44 (0.20) 3.84 (0.18) \0.001
Jump distance (mm) 3.04 (0.22) 2.92 (0.27) 0.252
Percent captured 4.10 (1.68) 22.85 (2.65) \0.001 6.50 (1.83) 25.39 (3.03) \0.001
Mean (SE). Results of Kruskal-Wallis One Way Analysis of Variance on Ranks (data not normally distributed). Acartia tonsa, comparative
values from Gilbert and Buskey (2005) and unpublished data
Table 2 Current speeds and turbulence at three speed settings for the
Flow Speed Experiment in the unidirectional flume and for the
oscillating flume
Unidirectional flume Oscillating flume
(cm s
speed (U)
(cm s
Mean speed
(cm s
3.2 0.089 11.6 7.8 0.077
9.2 0.288 24.6 15.8 0.35
26.7 1.203 40.4 24.2 1.01
Table 3 Current speeds and turbulent kinetic energy (TKE) in the
unidirectional flume at two speed settings for the Turbulence
Speed (cm s
) TKE (cm
3.6 3.2 -11.5 0.047 0.38 820
10.2 9.0 -12.5 0.18 1.76 970
Coral Reefs (2009) 28:475–488 481
Table 4;F
=28.49, P\0.001 for water speed;
=3.44, P=0.07 for species; F
=4.51, P=0.04
for species–water speed interaction; two-way ANOVA),
but under turbulent conditions, A. spinosa had a higher
foraging success than A. aspera (Fig. 5c; Table 4;F
27.77, P\0.001 for water speed; F
=6.94, P=0.01
for species; F
=0.68, P=0.41 for species–water
Foraging effort (approaches min
Mean attack distance (mm)Foraging success
3.2 9.2 26.7
Water speed (cm s-1)
Fig. 4 Unidirectional flume, Flow Speed Experiment. Foraging effort
a, attack distances band foraging success cfor Acanthemblemaria
aspera (light shading) and A. spinosa (dark shading).Error bars
indicate standard error
Foraging effort (approaches min-1)
Mean attack distance (mm)
Foraging success
3.4 cm s-1 9.6 cm s-1
Water speed
Smooth Turbulent Smooth Turbulent
Fig. 5 Unidirectional flume, Turbulence Experiment. Foraging effort
a, attack distances band foraging success cfor Acanthemblemaria
aspera (light shading) and A. spinosa (dark shading) at different flow
speeds and turbulence levels. Error bars indicate standard errors
482 Coral Reefs (2009) 28:475–488
speed interaction; two-way ANOVA). A. spinosa exhibited
higher foraging effort (Fig. 5a; F
=21.6, P\0.0001
for species; F
=0.22, P=0.64 for species–flow
regime interaction; F
=9.88, P=0.003 for flow spe-
cies–water speed interaction; three-way ANOVA) and
attack distances (Fig. 5b; F
=57.2, P\0.0001 for
species; F
=0.33, P=0.57 for species–flow regime
interaction; F
=22.1, P=0.011 for species–water
speed interaction; three-way ANOVA) than A. aspera in
combined smooth and turbulent flow as it did in smooth
flow alone (Figs. 4a and b).
Effect of oscillating flow on feeding efficiency
Foraging effort for both the species was the greatest at the
lowest mean flow speed (Fig. 6a). The flow speeds and
turbulence levels used in this experiment are presented in
Table 2. The mean values fall in the range of the pre-
dominant flow speeds experienced by the fish in nature
(Finelli et al. 2009), but the instantaneous top speeds
observed in the field (1.3 m s
) were greater than three
times the maximum speeds achieved in the flumes
(0.4 m s
For all flow conditions, A. spinosa exhibited a signifi-
cantly higher foraging effort than did A. aspera ranging
from 2.8 times higher at the lowest flow to 4.2 times higher
at the highest flow (Fig. 6a). Attack distances decreased
slightly (but significantly) with increasing water speed for
both the species, but there was no difference between the
species (Fig. 6b; F
=3.65, P=0.04 for water speed;
=2.71, P=0.11 for species; F
=0.38, P=0.69
for species–water speed interaction; two-way ANOVA).
Foraging success was significantly greater at the middle
speed for both the species and there was no difference
between the species (Fig. 6c; F
=3.73, P=0.04 for
water speed; F
=0.34, P=0.57 for species;
=0.51, P=0.61 for species–water speed interac-
tion; two-way ANOVA).
The times during an oscillating flow cycle when fish
extended maximally in their feeding strikes were measured
in seconds before or after the point of zero water flow
(Fig. 7). The standard deviations of these times were used
as a relative measure of the proportion of the wave cycle
that attacks occurred (Fig. 8b). At the lowest mean flow
speed (7.8 cm s
), A. spinosa approached over a greater
proportion of the wave cycle than did A. aspera (Fig. 7).
Table 4 Unidirectional flume, Flow Speed Experiment. Foraging
success for Acanthemblemaria aspera and A. spinosa feeding on
Acartia tonsa under different flow conditions
(cm s
Smooth flow Turbulent flow
Difference A.
3.4 0.28 0.27 -0.01 0.33 0.40 ?0.07
9.6 0.44 0.40 -0.04 0.57 0.64 ?0.07
7.8 15.8 24.2
Mean water speed (cm s-1)
Mean attack distance (mm) Foraging effort (approaches min-1)
Foraging success
Fig. 6 Oscillating flume. Foraging effort a, attack distances band
foraging success cfor Acanthemblemaria aspera (light shading) and
A. spinosa (dark shading) at three mean flow speeds (U
). Error bars
indicate standard error
Coral Reefs (2009) 28:475–488 483
As mean flow speed increased, A. aspera approached at a
fixed proportion of the wave cycle, whereas A. spinosa
approached over a reduced proportion of the wave cycle,
and at the highest flow speed, they did not differ in this
respect (Fig. 8b; F
=9.89, P=0.003 for species;
=2.28, P=0.12 for water speed; F
P=0.04 for species–water speed interaction; two-way
ANOVA). The lack of a significant effect of flow speed on
the proportion of the wave cycle used by A. aspera was
real, but due to the different patterns between the species
(significant interaction), it was artifactual for A. spinosa
(Fig. 8b). When tested alone, there was a significant
decrease in the amount of wave cycle used by A. spinosa
with increasing flow (F
=9.09, P=0.003, one-way
Both species foraged at higher instantaneous water
speeds as flow increased (Fig. 8c). A. spinosa attacked at
greater instantaneous water speeds than A. aspera at the
lowest flow speed (t =3.56, df =10, P=0.005), but
there was no difference between species at the highest flow
speed, which parallels the differences in standard devia-
tions (Fig. 8b). At the highest flow speed, the two species
utilized a similar proportion of the wave cycle (Fig. 8b)
and attacked at similar water speeds (Fig. 8c), but they also
exhibited the greatest difference in timing (Fig. 8a).
A. aspera attacked 0.18 s before slack flow at the highest
Time (s)
-1.00 -0.67 -0.33 0 0.33 0.67 1.00
Water seed (cm ) s-1
Water speed (cm s-1)
Fig. 7 Oscillating flume. Frequency of approaches (histograms) by
Acanthemblemaria aspera (top) and A. spinosa (bottom) over a half
‘wave’’ cycle (smooth curves) in the oscillating flume at 7.8 cm s
mean water speed (U
). Time is measured from slack point as
direction changes
Mean strike time (s)
Standard deviation (s)
Mean water speed at strike (cm s-1)
7.8 15.8 24.2
Mean water speed (cm s
Fig. 8 Oscillating flume. Mean strike time awhere positive times are
after and negative times are before the slack point (see Fig. 7),
standard deviation of strike times band mean absolute water speed at
strike time cfor Acanthemblemaria aspera (light shading) and
A. spinosa (dark shading) in the oscillatory flume at three mean flow
speeds (U
). Error bars indicate standard error
484 Coral Reefs (2009) 28:475–488
flow speed (Fig. 8a) whereas all other mean strike times
occurred after slack flow.
Water flow is critical to the distribution of coral reef fishes
by causing dislocations, increasing the energetic costs of
swimming, and, as shown here, by influencing their feeding
performance. Moreover, by modulating the ability of
planktivorous fishes to feed on benthic versus planktonic
food sources (Fig. 1), small scale variation in water flow
can alter spatial patterns of benthic-pelagic coupling. This
study demonstrates that water flow plays an important role
in ingestion rates of calanoid copepods by two species of
benthic tube blennies, and the patterns established here
suggest that the distribution of these species in the field is
related to their ability to meet metabolic energy require-
ments under varying flow conditions. For example,
consistent with its higher metabolic rate, A. spinosa had a
higher foraging effort and attack distance than A. aspera
under all conditions, but the two species did not differ in
foraging success across flow speeds. The two species
responded differently to changes in water flow: as water
speed increased from slow to medium, A. spinosa respon-
ded to the increased plankton delivery rate with a doubled
foraging effort, whereas the foraging effort of A. aspera
remained steady (Figs. 4a and 5a). These species-specific
feeding responses to water flow are consistent with the
spatial distribution of these blennies on coral reefs, and
suggest that critical processes modulated by planktivores
(e.g., benthic-pelagic coupling) are also modulated by
water flow.
A. spinosa attacked copepods at greater distances than
A. aspera in all cases in both the flumes (A. spinosa/
A. aspera mean ratio of 1.20, se =0.05, n=10) and in the
field (mean ratio of 1.67, se =0.14, n=4; Finelli et al.
2009). Because of its higher metabolic rate (Clarke 1992,
1999), A. spinosa has a higher food requirement. Its greater
success in capturing prey under turbulent conditions
(Fig. 5c, Clarke et al. 2005) suggests a greater capacity for
rapid adjustments, which also may make it less vulnerable
to predation during forays than is A. aspera. These features
may make it more likely to attack at considerably greater
distance, especially if the reward is large. In aquaria, both
species will attack adult Artemia sp. at distances as great as
30 to 40 cm and A. spinosa will go conspicuously farther
than A. aspera (R.D. Clarke pers. obs.). Inspection of the
video clips from the field showed that the greatest attack
distances (5 to 10 cm) were frequently associated with
larger prey. This may simply be a perception issue with
larger prey reaching some fixed minimum angular size in
the eye at a greater distance (Mussi et al. 2005), but could
also be the choice of a greater cost for a greater benefit
(Dill and Fraser 1984). The shorter attack distances for
A. spinosa in the flumes as compared to the field may
reflect the uniformly small prey provided (adult Acartia
tonsa ca. 840 lm long). The flumes were 100 mm wide,
enough to accommodate the distances observed for
A. spinosa in the field (Finelli et al. 2009).
The foraging effort for A. spinosa and A. aspera at the
two sites at Glover’s Reef was in the range of 0.3 to 0.6
approaches min
(Finelli et al. 2009). This is similar to
those reported for the same species on an artificial habitat
on a forereef in St. Croix, US Virgin Islands, where with
one exception, the mean values were in the range of 0.2 to
0.7 approaches min
(Clarke 1992). A slightly higher
range of 0.4 to 1.3 (mean 0.8) approaches min
recorded for 17 groups of 3 to 10 A. spinosa in Curac¸ao by
B. Luckhurst (pers. comm..). The foraging effort in the
flumes varied more widely than in the field, ranging from
0.13 to 6.9 approaches min
(Figs. 4a, 5a, and 6a). The
higher values in the flumes may be due to the high copepod
density used (5,000 individuals m
) compared with 11
daytime zooplankton densities of 25 to 2,500 individuals
reviewed in Heidelberg et al. (2004) and a maximum
of ca. 1,800 copepods m
in Fig. 2of Holzman et al.
(2005). Nevertheless, consistent with its greater food
requirement is the observation that the A. spinosa foraging
effort was higher than that of A. aspera under two condi-
tions on an artificial habitat in the field (Clarke 1992), and
under all conditions in both flumes in this study.
A. spinosa has been shown to feed primarily on plank-
tonic calanoid copepods, whereas A. aspera feeds primarily
on benthic harpacticoid copepods (Clarke 1999). In the
field, A. spinosa fed almost exclusively in the water col-
umn, whereas A. aspera ranged from exclusively benthic
feeding to almost exclusively feeding in the water column
(Finelli et al. 2009). In this respect, A. aspera somewhat
resembles Stegastes partitus (Pomacentridae), which feeds
on the substrate and ascends into the lower water column to
feed when a current is flowing (Stevenson 1972). The
stereotyped feeding location of A. spinosa as compared
with A. aspera is consistent with their attack speeds: the
former attacks at a fixed high speed under a wide range of
conditions, whereas the latter modulates its attack speed
according to conditions (Clarke et al. 2005; Waggett and
Buskey 2007). The correlation of planktonic feeding on the
reef with reduced water motion for A. aspera suggests a
limited capacity to function in energetic water conditions;
after all, plankton delivery would be higher with greater
flow making feeding in the water column more profitable
then, as shown by Stegastes partitus. Aside from a reduced
foraging success with increased water motion, the fish may
be more vulnerable to predation. Occasionally, in both field
and flume videotapes, a blenny extending beyond one body
Coral Reefs (2009) 28:475–488 485
length did not reenter the shelter directly. This seemed to
be due to an inability to maintain control under high flow
conditions and being swept away from the hole. The result
was a rapid maneuvering until the fish’s tail entered the
hole, a period during which the fish was both conspicuous
and unprotected. A need to minimize this vulnerable
behavior may play a role in the timing of attacks.
The reduced reaction distance and ability to evade pre-
dators that copepods experience at high flow speeds and
turbulence (Table 1, this study; Gilbert and Buskey 2005;
Robinson et al. 2007) results in greater foraging success by
both blennies under turbulent conditions (Fig. 5c). Under
prevailing conditions in the field, A. spinosa experienced
greater mean flow speeds and greater turbulence levels in
its microhabitat than did A. aspera (Finelli et al. 2009). It
also showed greater adaptation to these conditions by its
greater foraging success under turbulent conditions in
the unidirectional flume (Fig. 5c). In an earlier study,
A. spinosa also demonstrated higher foraging success than
A. aspera under turbulent conditions (Clarke et al. 2005).
Similarly, in the oscillating flume, A. spinosa attacked over
a greater proportion of the wave cycle than A. aspera at
mean flow speed 7.8 cm s
and 15.8 cm s
(Fig. 8b),
which translated to attacking when instantaneous flow
speeds were relatively high (Fig. 8c). This difference dis-
appeared at the highest mean flow speed, 24.2 cm s
where the two species attacked at the same instantaneous
water speeds, but A. aspera attacked before the slack time
and A. spinosa attacked after the slack time (Fig. 8a).
Given a mean return time of 0.11 s (Clarke et al. 2005),
this would place A. aspera at the shelter just at slack time,
maximizing its chance of a smooth reentry. This timing
may be more important to A. aspera if indeed it is a poorer
swimmer than A. spinosa.
It should be noted that the maximum mean flow speed in
the oscillating flume (24.2 cm s
) was below the maxi-
mum mean value recorded in the field (31 cm s
), where
8% of the mean flow speeds recorded on the forereef site
by Finelli et al. (2009) exceeded 24.2 cm s
. Neverthe-
less, foraging success did not differ between the species at
any speed (Fig. 6c), but note that the higher foraging effort
of A. spinosa resulted in its acquiring more food than
A. aspera. The declining foraging effort with increasing
mean flow speed (Fig. 6a) suggests that feeding would stop
at ca. 27 cm s
for A. aspera and ca. 30 cm s
A. spinosa (linear extrapolation to zero). The maximum
mean flow speed recorded on the forereef in this study was
31 cm s
(Finelli et al. 2009). The fish require some
minimum rate of food intake simply for maintenance; so
the mean flow speed to maintain a population must be well
below 30 cm s
Using the data gathered in this study, it is now possible
to estimate the energy intake of these two species of blenny
at different current speeds, and re-examine the conceptual
model presented in Fig. 1. A median blenny of either
species weighs approximately 80 mg and consumes a
minimum of 487.2 lLO
/d for A. aspera and 757.2 lLO
for A. spinosa (Clarke 1992,1999). Using an equivalency
of 4.63 kcal released per L O
consumed (Lovell 1998),
this gives a metabolic rate of 0.0016 cal/min for A. aspera
and 0.0025 cal/min for A. spinosa (dashed lines A and S in
Fig. 1).Ingestion rate would have to be higher because of
digestive inefficiency and energy needs not accounted for
by standard metabolic rate. Using the calculated values for
energy need, it can be seen that A. spinosa can just main-
tain itself at a water speed of 26.7 cm s
(Table 5, point D
in Fig. 1) and that A. aspera can only meet half its needs at
this water speed (given a copepod density of 5,000 cope-
pods m
). Both species can meet approximately 10 times
their needs at a water speed of 3.2 cm s
, but at
9.2 cm s
,A. spinosa can meet 25 times its need, whereas
A. aspera can meet nine times its need (Table 5). Since
blennies are diurnal feeders, they must ingest at a rate of
about two times their need during the day to meet their
24 h requirement. Note that the curve of ingestion rate
versus water speed for A. spinosa is dome-shaped, whereas
for A. aspera, it is not. This is consistent with the obser-
vations of Clarke et al. (2005) that A. aspera is the more
effective fish at capturing evasive prey in still water con-
ditions. It is possible that the curve for A. aspera would
also show a drop as water speeds approached zero, but
speeds below 3.2 cm s
were not tested in this study.
Table 5 Ingestion rates of Acanthemblemaria aspera and A. spinosa
at different flow speeds. Caloric values are based on the average
weights of the copepod Acartia tonsa (Tester and Turner 1988) and
the caloric value per unit weight of the copepod A. clausi (Kerabrum
1987). Proportion of need is the caloric intake divided by the standard
metabolic rate
Water Speed (cm min
) Ingestion rate
(copepods min
) (cal min
) Proportion of need
A. aspera A. spinosa A. aspera A. spinosa A. aspera A. spinosa
3.2 0.64 0.96 0.0166 0.0250 10.4 9.2
9.2 0.54 2.43 0.0140 0.0632 8.8 24.9
26.7 0.033 0.098 0.00086 0.00255 0.54 1.0
486 Coral Reefs (2009) 28:475–488
As shown by its (1) greater foraging distances, (2)
greater mean water speed during foraging, (3) greater
proportion of the wave cycle that it forages, and (4) greater
foraging success under turbulent conditions, A. spinosa is
more effective than A. aspera at capturing planktonic prey
under high velocity, turbulent water flow (Fig. 5). This
advantage may be achieved because its higher metabolic
rate could result in an ability to make quicker adjustments
during attacks and to swim more effectively in energetic
water flow, perhaps also reducing its vulnerability to pre-
dation. Its higher metabolic rate is sustained by its higher
foraging effort under all conditions. Thus, A. spinosa is
adapted to functioning under more challenging conditions
that are associated with greater food availability above the
benthic boundary layer and the food benefit outweighs the
metabolic cost (Table 5). Nevertheless, within the benthic
boundary layer, water motion increases with height above
the bottom (Shashar et al. 1996; Finelli et al. 2009); so the
horizontal axis in Fig. 1could also represent height above
the bottom. Only A. spinosa can occupy the locations to the
right of point C in Fig. 1, providing sole access to a portion
of the resource in short supply, holes in tall coral (Clarke
and Tyler 2003). Plankton availability close to the reef
surface is probably inadequate to sustain the higher meta-
bolic rate of A. spinosa, but A. aspera with its greater
flexibility can feed on the substrate and meet its lower food
requirements there.
The collapse of tall standing coral on Tague Bay Reef
in St. Croix, U.S. Virgin Islands, resulted in the almost
complete elimination of A. spinosa from that reef, while
A. aspera maintained its population density (Clarke 1996).
This study suggests that this shift in species composition
was related to the loss of topographically high shelters that
provided A. spinosa with access to energetic flow and
abundant plankton, while A. aspera was relatively unaf-
fected. Without high shelters, A. spinosa could not meet its
metabolic needs and was eliminated from the habitat. Such
results provide insight into how spatial complexity in
structure and flow enhances fish species diversity (Risk
1972; Molles 1978; Lirman 1999). Although the details
will be different for free-swimming planktivorous fishes,
differences in feeding behavior (in addition to vulnera-
bility to predators) at various flows may also play a role
in spatial separation between species and between size
classes within species (Hamner et al. 1988; Forrester
1991; Hobson 1991). Should this prove true, then the
reduced spatial complexity of degraded coral reefs can
reduce planktivore diversity and density via modulation
of feeding performance and access to resources. In turn,
alteration of planktivore distribution and feeding rates
can alter the rate and magnitude of benthic-pelagic
coupling, and reduce the productivity of the coral reef as
a whole.
Acknowledgements This work could not have been done without the
lab managers at the Middle Cay Field Station, Jon Clamp, Danny
Wesby, and Annick Cros, and the boatmen who navigated the complex
waters of Glover’s Reef, Elmar Avila, Cardinal Andrews, Faygon
Villanuevia, and Randolph Nunez. Field support was provided by Carly
Gaebe, Benjamin Kessler, Jora Ehrlich, and Eve Robinson. We wish to
thank D. Miller at the University of Delaware for suggesting the scotch
yoke design and R. Martin and R. Endsley at Lousiana Universities
Marine Consortium for flume construction. Analogue tapes were
converted to digital clips by Emma Mullaney, Yukari Kaito, Audrey
Black, Jora Ehrlich, Jacqui Beer, Olivia Peterson, and Meredith Gib-
bons. Douglas Brown provided free Tracker 1.3.2 software for video
motion analysis and Richard Lowrey provided the VassarStats website
on which the ANOVA analyses were done. This work was authorized
by a research permit from the Belize Fisheries Department, Ministry of
Agriculture and Fisheries and the University of Texas Institutional
Animal Care and Use Committee permit number 04113007. Funding
was provided by National Science Foundation Awards OCE-0324724,
OCE-0324413, and OCE-0324694/0715271. This is University of
Texas at Austin Marine Science Institute Contribution Number 1482
and Glovers Reef Research Station, Wildlife Conservation Society,
Publication Number 31.
Bellwood DR, Wainwright PC (2001) Locomotion in labrid fishes:
implications for habitat use and cross-shelf biogeography on the
Great Barrier Reef. Coral Reefs 20:139–150
Clarke RD (1992) Effects of microhabitat and metabolic rate on food
intake, growth and fecundity of two competing coral reef fishes.
Coral Reefs 11:199–205
Clarke RD (1996) Population shifts in two competing fish species on a
degrading coral reef. Mar Ecol Prog Ser 137:51–58
Clarke RD (1999) Diets and metabolic rates of four Caribbean tube
blennies, genus Acanthemblemaria (Teleostei: Chaenopsidae).
Bull Mar Sci 65:185–199
Clarke RD, Tyler JC (2003) Differential space utilization by male and
female spinyhead blennies, Acanthemblemaria spinosa (Teleo-
stei: Chaenapsidae). Copeia 2003:241–247
Clarke RD, Buskey EJ, Marsden KC (2005) Effects of water motion
and prey behavior on zooplankton capture by two coral reef
fishes. Mar Biol 146:1145–1155
Depczynski M, Bellwood DR (2005) Wave energy and spatial
variability in community structure of small cryptic coral reef
fishes. Mar Ecol Prog Ser 303:283–293
Dill LM, Fraser AHG (1984) Risk of predation and the feeding
behavior of juvenile coho salmon (Oncorhynchus kisutch).
Behav Ecol Sociobiol 16:65–71
Finelli CM, Clarke RD, Robinson HE, Buskey EJ (2009) Water flow
controls distribution and feeding behavior of two co-occurring
coral reef fishes: I Field measurements. Coral Reefs. doi:
Forrester GE (1991) Social rank, individual size, and group compo-
sition as determinants of food consumption by humbug
damselfish, Dascyllus aruanus. Anim Behav 42:701–711
Fulton CJ, Bellwood DR (2005) Wave-induced water motion and the
functional implications for coral reef fish assemblages. Limnol
Oceanogr 50:255–264
Fulton CJ, Bellwood DR, Wainwright PC (2005) Wave energy and
swimming performance shape coral reef fish assemblages. Proc
R Soc B 272:827–832
Gilbert OM, Buskey EJ (2005) Turbulence decreases the hydrody-
namic predator sensing ability of the calanoid copepod Acartia
tonsa. J Plankton Res 27:1067–1071
Coral Reefs (2009) 28:475–488 487
Gourlay MR, Colleter G (2005) Wave-generated flow on coral reefs–
an analysis for two-dimensional horizontal reef-tops with steep
faces. Coast Eng 52:353–387
Hamner WM, Jones MS, Carleton JH, Hauri IR, Williams DM (1988)
Zooplankton, planktivorous fish, and water currents on a
windward reef face: Great Barrier Reef, Australia. Bull Mar
Sci 42:459–479
Heidelberg KB, Sebens KP, Purcell JE (2004) Composition and
sources of near reef zooplankton on a Jamaican forereef along
with implications for coral feeding. Coral Reefs 23:263–276
Hobson ES (1991) Trophic relationships of fishes specialized to feed
on zooplankters above coral reefs. In: Sale PF (ed) The ecology
of fishes on coral reefs. Academic Press, San Diego, pp 69–95
Holbrook SJ, Brooks AJ, Schmitt RJ, Stewart HL (2008) Effects of
sheltering fish on growth of their host corals. Mar Biol 155:
Holzman R, Reidenbach MA, Monismith SG, Koseff JR, Genin A
(2005) Near-bottom depletion of zooplankton over a coral reef
II: relationships with zooplankton swimming ability. Coral Reefs
Johansen JL, Bellwood DR, Fulton CJ (2008) Coral reef fishes exploit
flow refuges in high-flow habitats. Mar Ecol Prog Ser 360:
Kench PS (1998) Physical processes in an Indian Ocean atoll. Coral
Reefs 17:155–168
Kerabrum P (1987) Composition chimique e
´mentaire (C, H, N) et
´quivalent e
´nergie d’Acartia clausi (Crustacea: Copepoda),
`ce importante dans la bioe
´tique des e
ˆtiers de Me
´e nord-occidentale. Mar Biol 95:115–121
Kiflawi M, Genin A (1997) Prey flux manipulation and the feeding
rates of reef-dwelling fish. Ecology 78:1062–1077
Kotrschal K, Lindquist DG (1986) The feeding apparatus in four
Pacific tube blennies (Teleostei: Chaenopsidae): lack of eco-
morphological divergence in syntopic species. PSZNI Mar Ecol
Liberman T, Genin A, Loya Y (1995) Effects on growth and
reproduction of the coral Stylophora pistillata by the mutualistic
damselfish Dascyllus marginatus. Mar Biol 121:741–746
Lindquist DG (1985) Depth zonation, microhabitat, and morphology
of three species of Acanthemblemaria (Pisces: Blennioidea) in
the Gulf of California, Mexico. PSZNI Mar Ecol 6:329–344
Lirman D (1999) Reef fish communities associated with Acropora
palmata: relationships to benthic attributes. Bull Mar Sci
Lithgoe JN, Partridge JC (1989) Visual pigments and the acquisition
of visual information. J Exp Biol 146:1–20
Lovell T (1998) Nutrition and feeding of fish, 2nd edn. Kluwer,
Norwell, MA
MacKenzie BR, Miller TJ, Cyr S, Leggett WC (1994) Evidence for a
dome-shaped relationship between turbulence and larval fish
ingestion rates. Limnol Oceanogr 39:1790–1799
Madin JS, Black KP, Connolly SR (2006) Scaling water motion on
coral reefs: from regional to organismal scales. Coral Reefs
Molles CM Jr (1978) Fish species diversity on model and natural reef
patches: experimental insular biogeography. Ecol Monogr
Mussi M, McFarland WN, Domenici P (2005) Visual cues eliciting
the feeding reaction of a planktivorous fish swimming in a
current. J Exp Biol 208:831–842
Porat D, Chadwick-Furman NE (2005) Effects of anemonefish on
giant sea anemones: ammonium uptake, zooxanthella content
and tissure regeneration. Mar Freshw Behav Physiol 38:43–51
Risk MJ (1972) Fish diversity on a coral reef of the Virgin Islands.
Atoll Res Bull 153:1–6
Robinson HE, Finelli CM, Buskey EJ (2007) The turbulent lives of
copepods: effects of water flow over a coral reef on their ability
to detect and evade predators. Mar Ecol Prog Ser 349:171–181
Shashar N, Kinane S, Jokiel PL, Patterson MR (1996) Hydrome-
chanical boundary layers over a coral reef. J Exp Mar Biol Ecol
Stevenson RAJ (1972) Regulation of feeding behavior of the bicolor
damselfish (Eupomacentrus partitus Poey) by environmental
factors. In: Winn HE, Olla BL, Burger J (eds) Behavior of
marine animals: perspectives in research. Plenum, New York,
pp 278–302
Tester PA, Turner JT (1988) Comparative carbon-specific ingestion
rates of phytoplankton by Acartia tonsa,Centropages velificatus
and Eucalanus pileatus grazing on natural phytoplankton
assemblages in the plume of the Mississippi River (northern
Gulf of Mexico continental shelf). Hydrobiologia 167(168):
Toomas L (2001) Investigation of turbulence in a plunging breaking
wave. Proc Estonian Acad Sci Eng 7:58–78
Waggett RJ, Buskey EJ (2007) Calanoid copepod escape behavior in
response to a visual predator. Mar Biol 150:599–607
White JW, Warner RR (2007) Behavioral and energetic costs of group
membership in a coral reef fish. Oecologia 154:423–433
Yahel R, Yahel G, Genin A (2005) Near-bottom depletion of
zooplankton over coral reefs: I: diurnal dynamics and size
distribution. Coral Reefs 24:75–85
Young IR (1989) Wave transformation over coral reefs. J Geophys
Res 94:9779–9789
488 Coral Reefs (2009) 28:475–488
... However, the adaptive distinctions that accompany niche partitioning are not necessarily reflected in outward anatomical form. Behavioral differences, commonly paired with divergent physiological adaptations, can also have implications for resource use, constituting important niche determinants among morphologically and functionally-similar organisms (Schmitt and Coyer 1982;Hartney 1989;Clarke et al. 2009). The conspicuous absence of any defining morphological characters between the species studied here seems to suggest that the disparate flow environments exploited by A. goreensis and A. vulpes may thus be related to differences in behavior and/or internal physiology. ...
... Interspecific differences in metabolism may also have contributed to observed interspecific contrasts in otolith δ 13 C. Assuming the species display equivalent δ 13 C fractionation and have analogous dietary inputs, the isotopically-lighter otolith δ 13 C values of A. goreensis may be interpreted to suggest that this species maintains a higher metabolic rate (Kalish 1991;Høie et al. 2003), consistent with the greater energetic demands required by the comparatively high-flow or turbulent habitats it occupies (Enders et al. 2003;Roche et al. 2014). Similar discrepancies in speciesspecific metabolism or activity level have been linked with differential microhabitat and resource use among other sympatric congeners, and may represent adaptations that help to balance habitat-specific energetic costs and resource availability (Hartney 1989;Clarke et al. 2005Clarke et al. , 2009. The more widely ranging δ 13 C oto values observed for A. goreensis suggest that this species exploits a greater assortment of resources or microhabitats, in accordance with its broader distribution among sampling stations and the notably more heterogenous hydrodynamic regimes it occurred in. ...
Full-text available
We employed numerical wave models, GIS, and stable isotope analyses of otolith material to identify interspecific differences in habitat and resource use among juveniles of two sympatric and morphologically indistinct bonefishes, A. goreensis and A. vulpes in littoral zones of The Bahamas. Both species occurred in similar water temperatures; however, A. goreensis juveniles occupied habitats characterized by greater wave-driven flow velocities and closer proximity to coral reefs than A. vulpes. Likewise, A. goreensis was present across a broader range of flow environments and sampling stations than A. vulpes, which was typically confined to sheltered, low-flow habitats. The results of stable isotope analyses were consistent with the species’ relationships with environmental parameters, providing support for differential habitat and/or resource utilization. Otolith δ¹⁸O did not differ significantly between species, suggesting they experience comparable thermal regimes. However, δ¹³C varied substantially, with the otoliths of A. goreensis depleted in ¹³C relative to A. vulpes by approximately 1‰, potentially signifying a greater reliance on pelagic carbon sources by the former, in agreement with observed distinctions in habitat use. In linear models, otolith δ¹³C was negatively correlated with ambient flow velocity and positively related to distance from coral reef habitats, and these relationships did not vary across species. After accounting for the effects of these variables, species-specific differences in otolith δ¹³C remained, indicating that other unknown factors contributed to the observed disparities. Collectively, our findings suggest that niche partitioning between A. goreensis and A. vulpes is likely mediated by their differential abilities to compete across various flow environments, likely as a result of divergent behavioral and/or physiological adaptation.
... station-holding). By inducing rheotaxis in a modified respirometer (SWMR; Fig. 1), we were able to induce swimming at wave periods (3.3-10 s) and velocities (0-29 cm s −1 ) within the range of flows measured on coral reefs Lowe et al., 2005;Denny, 2006;Madin et al., 2006;Clarke et al., 2009). The metabolic rate of fishes swimming in this oscillatory pattern of wave surge was found to increase as both frequency and amplitude of wave action increased, indicating higher metabolic costs to station-hold as wave action increases (Fig. 5). ...
Full-text available
Oxygen consumption rates were measured for coral reef fishes during swimming in a bidirectional, oscillatory pattern to simulate station-holding in wave-induced, shallow water flows. For all species examined, increases in wave intensity, as simulated by increases in frequency and amplitude of oscillation, yielded increased metabolic rates and net costs of swimming (NCOS; swimming metabolic rate minus standard metabolic rate). Comparing species with different swimming modes, the caudal-fin swimming Kuhlia spp. (Kuhliidae) and simultaneous pectoral-caudal fin swimming Amphiprion ocellaris (Pomacentridae) turned around to face the direction of swimming most of the time, while the median-paired fin (MPF) swimmers, the pectoral -fin swimming Ctenochaetus strigosus (Acanthuridae) and dorsal-anal fin swimming Sufflamen bursa (Balistidae), more frequently swam in reverse for one half of the oscillation to avoid turning. Contrary to expectations, the body-caudal fin (BCF) swimming Kuhlia spp. had the lowest overall NCOS in the oscillatory swimming regime compared with the MPF swimmers. However, when examining the effect of increasing frequency of oscillation at similar average velocities, the Kuhlia spp. showed a 24% increase in NCOS with a 50% increase in direction changes and accelerations. The two strict MPF swimmers had lower increases on average, suggestive of reduced added costs with increasing frequency of direction changes with this swimming mode. Further studies are needed on the costs of unsteady swimming to determine if these differences can explain the observed prevalence of fishes using the MPF, pectoral-fin swimming mode in reef habitats exposed to high, wave-surge induced water flows.
... The planktivorous species dominated natural reef sites in this study is surprisingly contrast to the earlier reports on fish trophic structure from shipwreck and natural reefs (Rilov and Benayahu, 2000;Simon et al., 2013). These species prefer waters with high plankton availability and the abundance of plankton is influenced by primary productivity, incidence of solar radiation, wave action and current strength (Clarke et al., 2009). The shallow natural reef sites near the surface have reduced planktivore biomass on account of resistance to water movements, reduced currents and plankton movement. ...
Aim: The Objective of the study was to test the hypothesis that submerged shipwrecks along western coast of India can function as artificial fish habitats with trophic structure of fish assemblages similar to natural rocky reef habitats. Methodology: The fish assemblages on shipwreck and natural reefs were analysed from September 2013 to December 2016 using diver assisted underwater visual census (UVC). Data collected on fish assemblages were used to compare the fish trophic structure between shipwreck and natural reefs using diversity indices, permutational multivariate analysis of variance and similarity percentage analysis. Results: Fish abundance was two times greater on shipwreck than on natural reefs and the major trophic guilds on shipwreck were omnivores and carnivores. Multivariate analyses showed a distinct fish trophic structure on shipwreck from that of natural reefs and certain fish species differentiated the trophic structure of shipwreck from natural reefs. The high densities of Pomadasys furcatus, P. guoraca, Pempheris multiradiata, Lutjanus indicus, L. fulvus, Ostorhinchus compressus,, Epinephelus coioides, E. erythrurus, Monodactylus argenteus and Heniochus singularis contributed significantly to the average dissimilarity for fish assemblages in shipwreck from natural reefs. This has helped to differentiate the fish communities on shipwreck from natural reefs. Interpretation: This study showed that shipwreck and natural reefs differ significantly in fish trophic structure. Fish species richness and abundance were significantly higher on shipwreck than natural reefs. Therefore, the Greater species richness, abundance and diversity of fish communities on shipwreck could be a result of unique, complex and heterogeneous habitat features exist within the wreck.
... In rivers and streams, drift-feeding fishes maximize their net energy gain by selecting microhabitats that provide velocity refuges (e.g., rocks and wood debris) to reduce swimming costs in close proximity to faster flows that supply prey (Hughes and Dill 1990;Blanchet et al. 2008;Fausch 2014). Planktivorous fishes inhabiting marine reefs benefit similarly from prey transport by currents and waves (Anderson and Sabado 1995;Clarke et al. 2009;Finelli et al. 2009) and, more generally, by intermediate levels of turbulence in the plankton that are high enough to enhance encounter rates but low enough to not reduce pursuit and capture of prey (MacKenzie et al. 1994). In lakes, waves are known to enhance fish feeding by suspending benthic invertebrates into the water column (Gabel et al. 2011). ...
... This has led to the demonstration that the C1 copepodite has an improved escape ability and predator avoidance compared with the N6 nauplius (Buskey, 1994;Bradley, 2009), but this has not been tested for visually hunting, natural predators. Water motion should also be considered in these predatorprey interactions since turbulence impacts the ability of adult copepods to detect hydrodynamic disturbances and also affects capture success by visual predators (Gilbert and Buskey, 2005;Waggett and Buskey, 2007;Clarke et al., 2009). ...
Full-text available
Escape success of copepod nauplii (N5–6) of the copepod Acartia tonsa was significantly lower than that of copepodites (C1–2) when exposed to a coral reef fish Acanthemblemaria paula, but no difference was observed when exposed to the seahorse Hippocampus zosterae. The addition of low level turbulence inhibited feeding by the seahorse but did not affect overall capture success in A. paula, although a shift in predator behavior occurred. This study highlights the importance of considering predatory mode and behavior in the context of planktonic predator–prey dynamics.
Full-text available
Feeding by zooplanktivorous fish depends on their foraging movements and the flux of prey to which they are exposed. While prey flux is a linear function of zooplankton density and flow speed, those two factors are expected to contribute differently to fish movements. Our objective was to determine effects of these factors for garden eels, stationary fish that feed while anchored to the sandy bottom by keeping the posterior parts of their bodies inside a burrow. Using a custom-made flume with a sandy bottom, we quantified effects of prey density and flow speed on feeding rates by spotted garden eels (Heteroconger hassi). Feeding rates increased linearly with prey density. However, feeding rates did not show a linear relationship with flow speed and decreased at 0.25 m s−1. Using label-free tracking of body points and three-dimensional movement analysis, we found that the reduction in feeding rates was related to modulation of the eel's movements, whereby the expected increase of energy expenditure was avoided by reducing exposure and drag. No effects of flow speed on strike speed, reactive distance, or vectorial dynamic body acceleration (VeDBA) were found. A foraging model based on the body length extended from the burrow showed correspondence with observations. These findings suggest that due to their unique foraging mode, garden eels can occupy self-made burrows in exposed shelter-free sandy bottoms where they can effectively feed on drifting zooplankton.
Full-text available
Ecological niches hold critical information concerning the eco-evolutionary dynamics that govern biodiversity and abundance patterns. Cryptobenthic reef fishes account for approximately half of all reef fish species and are an abundant and important group on coral reefs worldwide. Yet, due to their small size and inconspicuous lifestyles, relatively little is known about the ecological niches of most cryptobenthic species. Here, we use gut content DNA metabarcoding to determine dietary niche overlap and prey richness in four sympatric species of cryptobenthic reef fishes in two genera (Acanthemblemaria aspera, A. spinosa, Enneanectes altivelis, and E. matador). Furthermore, we test whether dietary differentiation corresponds with differences in species distribution patterns across twelve sites on the Mesoamerican Barrier Reef in Belize. Our approach reveals dietary partitioning among the four species, which is further supported by low edge density and high modularity in the resulting trophic network. A. spinosa and E. matador consume a significantly higher richness of prey items than their congeners. This result corresponds with non-random distributions and co-occurrence patterns in both species pairs: the two high prey richness species (A. spinosa and E. matador) co-occur more frequently than predicted by chance, but they are exclusive to exposed forereef sites with high wave action. In contrast, their congeners occur across exposed forereef and sheltered backreef sites, but they do not increase in numbers at sheltered sites. Our findings suggest that A. spinosa and E. matador monopolize a wide variety of prey in exposed habitats, but they are unable to meet the energetic demands of their adaptation to high-flow habitats in sheltered areas, possibly due to lower prey availability. This, in turn, indicates strong ecological differentiation among closely related species of cryptobenthic fishes, driven by links between diet, physiology, prey availability, and wave exposure.
Full-text available
The distribution, abundance, and habitat preferences of the spinyhead blenny, Acanthemblemaria spinosa (Perciformes, Blennioidei, Chaenopsidae), were studied on coral reefs along the leeward side of Curaçao, southern Caribbean. The blennies inhabited small holes inside coral, which predominantly consisted of calcareous tubes constructed by coral-associated serpulid worms of the species Spirobranchus giganteus. About 50 % of the fish inhabited holes in dead coral, and the rest had their holes in live corals of eight species. The fishes showed a clustered distribution pattern and their abundance was higher at shallow depths (5 and 10 m) than at 15 m. Although males generally had a larger body size than females and needed larger holes for shelter and guarding eggs, no sexual dominance in hole selection was found. The position of the holes varied in elevation height above the reef floor, which showed a positive correlation with fish size.
Benthic suspension feeders are a primary conduit for the transfer of carbon from the water column to the benthos. As such, factors that influence their feeding require mechanistic study and quantification. In this flume experiment, the clearance rate of the solitary ascidian Styela plicata varied as a function of flow speed, body size, and particle diameter. At all flow speeds and body sizes tested, clearance rates increased directly with particle diameter to similar to 10 mu m. As particle diameter increased further to similar to 32 mu m, clearance remained constant or declined depending on flow speed. At 3 and 22 cm s(-1), clearance remained constant with particle diameter > 10 mu m. At 14 cm s(-1), clearance sharply declined as particle diameter increased above 10 mu m. In general, clearance rates increased with body size across all particle diameters. However, allometric exponents relating clearance rate to body size, which ranged from 0.28 to 0.62, were lower than expected (similar to 0.67) due to the confounding effects of water flow, to which clearance rate responded in a non-linear fashion. We fit our measurements to a 3-dimensional surface that relates clearance rate, body size, and flow speed. These surface fits show that clearance of small particles (< 10 mu m diameter) was maximal at intermediate flow speeds (similar to 12 cm s(-1)) and decreased at both faster and slower flow speeds. This 'unimodal' response is consistent with predictions of suspension-feeding theory. In contrast, clearance of large particles (> 10 mu m diameter) decreased steadily as flow speed increased from 3 to 22 cm s(-1).
Full-text available
The heterogeneous distribution of water flow over structurally complex environments, such as coral reefs, may play an important role in the interactions between copepods and planktivorous fish by interfering with the copepods' ability to detect and evade predators. The escape response and capture rates of the copepod Acartia tonsa were examined in laboratory flumes that created both unidirectional and oscillatory flow conditions similar to those found near coral reefs. Two turbulent regimes were produced in each flume: 'smooth' flow was formed using a grid collimator and 'rough' flow was generated by placing a branched coral skeleton upstream of the flume's working section. A predator was simulated by a fixed siphon to generate a stimulatory flow field. Copepod detection of the siphon was measured as the distance from the siphon tip to where an escape response was initiated. This reactive distance remained the same in low-flow conditions as in still water, but was reduced by 25% at higher flow speeds, indicating a decline in the copepods' ability to detect velocity gradients formed by the siphon. Rough turbulence regimes intensified the effect of current speeds, resulting in an even shorter copepod reactive distance. Capture rates of copepods by the siphon increased with current speed, wave motion, and in rough flow, while the capture rates of non-evasive prey, Artemia nauplii, did not vary with flume conditions. The differences in capture rates between evasive and non-evasive prey suggest that behavioral shifts in copepod escape thresholds may account for increases in predation by reef-dwelling fishes observed in hydrodynamically complex coral environments.
Full-text available
I monitored the population~ of 2 fishes on a coral reef in St. Croix, U. S. Virgin Islands, from 1980 to 1995 to determine if they could track changes in their habitat. Starting in the mid-1970s. white band disease killed all the elkhorn corals Acropora palmata, providing access to a variety of boring organisms that create the cavities inhabited by the spinyhead blenny Acanthemblemaria spinosa and the roughhead blenny A aspera As the corals became more heavily bored, they were weakened and collapsed, providing more of the low habitat occupied by roughheads and less of the high h a b ~ t a t occupied by spinyheads. This structural change was generally gradual but Hurricane Hugo destroyed almost all the remain~ng standing dead corals in 1989 Both specles increased in density until 1989 when the sudden reduct~on of high habitat caused a decrease in spinyhead population density. In contrast, roughheads continued their population growth until 1991 after which they also declined. Although previous work has shown that spinyheads can displace roughheads from preferred habitat, spinyheads did not increase their numbers close to the reef surface as the high corals collapsed. I hypothesize that their inflexible minimum height is due to their previously demonstrated higher metabolic rate which constrains them to the higher locations where planktonic food is more abundant. Consequently, the species mix shifted from one of spinyheads with extremely rare roughheads in 1980 to one dominated by roughheads w t h few splnyheads in 1995. This precise tracking of a changing environment suggests resource limitation and is counter to the widespread view that coral reef fish assemblages are largely unsaturated systems consisting of open, 1-ecruitment-limited populations. The fishes studied here may be atypical, however, and it is suggested that more studies with a mechanistic approach may prov~de insight into reef fish assemblage structure.
Aquaculture is now recognized as a viable and profitable enterprise worldwide. As aquaculture technology has evolved, the push toward higher yields and faster growth has involved the enhancement or replacement of natural foods with prepared diets. In many aquaculture operations today, feed accounts for more than one-half the variable operating cost. Therefore, knowledge of nutrition and practical feeding of fish is essential to successful aquaculture. This book is not written exclusively for scientists but also for students, practicing nutritionists, and aquaculturists. It covers the known nutrient requirements and deficiency effects for different fishes, and digestion and metabolism of nutrients and energy. It discusses nutrient sources and preparation of practical and research feeds. It gives directions for conducting fish nutrition and feeding experiments. Feeding practices for salmonids, channel catfish, tilapias, shrimps and hybrid striped bass are presented. Since the first edition of this book was printed, the National Research Council of the National Academy of Sciences has revised the nutrient requirements for fish. These revisions are in the present edition. Other additions to this revised edition are chapters on nutrition and fish health, and bioavailability of nutrients. Each original chapter has been meticulously revised and updated with new information. Aquaculture is a dynamic area and new technologies are being introduced continuously; therefore, some of the material discussed in this revised edition may become obsolete quickly. Nonetheless, the material presented has been thoughtfully selected and updated to make it of maximum use to persons whose interests range from general aquaculture to animal nutrition to feed manufacture.
Studies of food chains and webs on reefs have been carried out by collecting and observing animals, identifying and quantifying the kinds of food they have eaten, and thereby determining their ecological relationships. Emery (1968), Hobson (1968), Randall (1967), Hiatt and Strasburg (1960), and others showed that certain species of fishes feed predominantly on particular kinds of animals or plants and thus have established pathways through which food, and energy, “flows.”
Nutrients must be provided in appropriate amounts and in forms that are biologically usable for optimum performance by the animal. Therefore it is as important to know the bioavailability of the nutrient as the dietary requirement. A respectable amount of data is available on digestibility of gross energy and crude protein in commercial ingredients used in fish feeds. There is, however, much less information on bioavailability of vitamins, minerals and amino acids from various natural and synthetic sources. In many cases, assumed availability values for nutrients are used to formulate fish feeds which are probably far from accurate. Examination of data presently available supports this contention.
Fish communities associated with model and natural reef patches were studied to determine the effects of heterogeneity, area, and isolation on the number of fish species in these systems. The effects of season and succession on the structure of these communities were also examined. Species turnover was estimated to determine if numbers of fish species on small reefs can be viewed as a balance between immigration and local extinction. Interspace size diversity, one facet of reef heterogeneity, was not positively correlated with the number of fish species on either model or natural reefs. In addition, interspace size diversity had no significant effects on species composition. In most of the cases examined reef height, another aspect of reef heterogeneity, was positively correlated with the number of fish species and species diversity, H', on natural reefs. Observations of selected fish species inhabiting model reefs support the hypothesis that vertical zonation is a means of resource partitioning in these fish communities. A negative correlation was found between reef isolation and number of fish species on patch reefs. This correlation was strongest during periods when fish population sizes were lowest. Season had more of an effect on the structure of fish communities on model reefs than did succession. Immigration and local extinction of fish species occurred on model reefs throughout the study. Patterns of immigration and extinction approximated the predictions of the MacArthur-Wilson equilibrium model of insular zoogeography when species turnover was highest.
The velocity field created by a plunging breaking wave on a smooth bottom with slope 1 : 17 was studied experimentally in a wave flume. Laser Doppler anemometry was used to investigate the flow field above the bottom boundary and below the trough level of the wave. Turbulence intensities, Reynolds stresses, and turbulent kinetic energy were examined. The results show that large-scale motions dominate in turbulence under the plunging breaker. The flow has characteristic features of an outer surf zone. It is found that turbulent quantities in the zone close to the bottom depend on the nature of the flow acceleration. During the deceleration phase, all turbulent quantities reach their maximum values. In the layers close to the wave trough, turbulent quantities depend on the wave parameters. Turbulent kinetic energy reaches its maximum value under the wave crest and decreases rapidly to a constant value under the wave trough. Turbulence is generated on the surface during the breaking process and it diffuses towards the bottom. The energy level first decreases downward and then increases again close to the bottom due to the bottom boundary layer turbulence. Kinetic energy is transported landward in the upper layers of the flow and seaward near the bottom.
Carbon, hydrogen and nitrogen content were investigated in natural populations of the crustacean copepod Acartia clausi collected during 1980–1982 from different areas near Marseilles, France — the coastal area of Cortiou which receives waste waters from the Marseilles Conurbation; the eutrophic and polluted Golfe de Fos, with its open basins; and the offshore areas of the Marseilles' Gulf. Energy equivalents were determined in order to examine the ecological role of A. clausi from a bioenergetic point of view. Carbon content ranged from 38.9% of body dry weight (DW) in the open sea off Marseilles to 45.5% of DW in an open basin at Fos in winter, nitrogen content from 10.4% of DW in the open sea off Marseilles in winter to 13.6% of DW in the open basins at Fos in summer. These values agree with the ranges reported in the literature for non-lipidic copepods. Carbon to nitrogen ratios ranged between 3.22 and 3.80. In the open basins at Fos, the mean value in summer (3.30) was significantly different from that in winter (3.70). In the Cortiou area, no seasonal differences emerged. For the three areas, energy equivalents ranged from 4.28 to 5.00 cal mg-1 DW. Highest values were recorded for the populations of the Cortiou area and of the open basins at Fos, i.e., from environments where particulate organic matter is abundant and relatively invariable. In these populations, the energy equivalent per unit dry weight did not change significantly throughout the year. The values recorded agreed with the means reported in the literature for non-lipidic species, but were far below those recorded for species which are able to store lipids. A. clausi is one of the most abundant copepod species throughout the year in the plankton of these coastal areas, and its steady energy content, with little seasonal variation, makes it an important element in the bioenergetics of the northwest Mediterranean coastal ecosystems.