312

Good eaters, poor swimmers: compromises in larval form

Richard R. Strathmann1,* and Daniel Grünbaumy
*Friday Harbor Laboratories, University of Washington, 620 University Road, Friday Harbor, WA 98250, USA;
ySchool of Oceanography, Box 357940, University of Washington, Seattle, WA 98195-7940, USA

Synopsis Compromises between swimming and feeding affect larval form and behavior. Two hypotheses, with supporting
examples, illustrate these feeding-swimming trade-offs. (1) Extension of ciliated bands into long loops increases maximum
clearance rates in feeding but can decrease stability of swimming in shear flows. A hydromechanical model of swimming by
ciliated bands on arms indicates that morphologies with high performance in swimming speed and weight-carrying ability in
still water differ from morphologies conferring high stability to external disturbances such as shear flows. Instability includes
movement across flow lines from upwelling to downwelling water in vertical shear. Thus a hypothesis for the high arm
elevation angles of sea urchin larvae, which reduce speed in still water, is that they reduce a downward bias imposed by the
vertical shear in turbulence. Observations of sea urchin larvae in vertical shear and comparisons among brittle star larvae are
consistent with the performance trade-offs predicted by the model. (2) Structures and behaviors that reduce swimming speed
can enhance filtering for feeding. In the opposed-band feeding mechanisms of veligers and many trochophores, cilia push
water to swim but movement of cilia relative to the water occurs when cilia overtake and capture particles. Features that may
increase clearance rates at the expense of speed and weight capacity include structures that increase drag or body weight and a
ciliary band that beats in opposition to the feeding-swimming current. Larval feeding mechanisms inherited from distant
ancestors result in different swimming-feeding trade-offs. The different trade-offs further diversify larval form and behavior.

Introduction (Emlet and Strathmann 1985; Fenchel and Ockelmann

Swimming helps planktotrophic larvae move beyond
water that they have already processed for food and
may help them find aggregations of prey. Swimming
and feeding are different processes, however, and there
are trade-offs between performance in clearing water of
food and performance in swimming. These trade-offs
differ among larvae and among environments for the
same larvae. Swimming-feeding trade-offs may under-
lie divergence in form and behavior and differences
in preferred habitat.
Here we discuss 2 hypotheses on compromises
between swimming and feeding. One hypothesis is
that the long arms and lobes that increase maximum
clearance rates impair stability of swimming in shear
(Grünbaum and Strathmann 2003). This hypothesis is
examined in particular for the pluteus larvae of ech-
inoids (sea urchins) and ophiuroids (brittle stars), but
may apply to any larvae with long arms or lobes. The
trade-off involves the overall form of the larvae and
may explain the reduction or absence of arms or lobes
in related non-feeding larvae. The other hypothesis
is that there is a compromise between pushing water
to swim and creating shear around cilia or setae to feed
2002). This hypothesis is examined in particular for
larvae that feed with opposed prototrochal and meta-
trochal bands (as do many mollusc and annelid larvae)
but may apply to other arrays of cilia or setae that both
propel the animal and filter particles from suspension.
Larval arms or lobes: stability versus
speed and weight capacity
Arms and clearance rates
Diverse larvae capture small planktonic food with
bands of cilia. This kind of food is often at low con-
centrations. To achieve a high rate of ingestion, the
larva must clear water of suspended food at a high
rate. Clearance rate is the volume of water cleared of
food particles per time. The maximum clearance rate
depends to a great extent on the length of ciliary bands.
One way to support long ciliary bands on a small larval
body is to extend the band on narrow arms, tentacles,
or lobes. Ciliary bands on arms or lobes are character-
istic of the pluteus larvae of sea urchins and brittle
stars, the actinotroch larvae of phoronid worms, and
the feeding larvae of brachiopods. They occur in some

From the symposium “Complex Life-Histories in Marine Benthic Invertebrates: A Symposium in Memory of Larry McEdward” presented at the
annual meeting of the Society for Integrative and Comparative Biology, January 4–8, 2005, at San Diego, CA.
1 E-mail: rrstrath@u.washington.edu

Integrative and Comparative Biology, volume 46, number 3, pp. 312–322

doi:10.1093/icb/icj031
Advance Access publication March 29, 2006
Ó The Society for Integrative and Comparative Biology 2006. All rights reserved. For permissions, please email: journals.permissions@oxfordjournals.org.
Good eaters, poor swimmers 313

gastropod molluscs, and annelid worms. For all these
larval forms, observations of feeding mechanisms
indicate that the bands along the arms or lobes are
effective in the capture of food particles (Strathmann
1987). The ciliated arms of the actinula larvae of
trachyline medusae may also capture prey. For the
echinoderm larvae, maximum clearance rates are pro-
portional to length of the ciliary band (Strathmann
1971; Hart 1996). Feeding currents indicate a similar
proportionality for other larvae with ciliary bands on
arms or lobes.
Arms and stability
Long arms or lobes might compromise swimming in
2 ways. First, body extensions might compromise sta-
bility in shear. Long body dimensions are exposed to
greater shear that could change orientation. Second,
anteriorly directed arms could shift the center of grav-
ity relative to the center of buoyancy. Most larvae are
passively stable so that they swim upward in still water
unless they actively change direction. Dense structures
positioned posteriorly in larvae confer passive vertical
orientation because their center of gravity lies below
their center of buoyancy (Fig. 1C) (for example,
Pennington and Strathmann 1990; Mogami and others
2001; Fenchel and Ockelmann 2002). In plutei, the
anteriorly directed arms (Fig. 1A and B) are associated
with a posterior counterweight that maintains a stable
orientation (Pennington and Strathmann 1990). The
skeleton of arms and the posterior counter weight
result in a larval density of about 1.2–1.25 g cm
3 in
plutei of a sand dollar, substantially greater than that of
seawater (Pennington and Emlet 1986).

0.1

Z β
0 α
−0.2
−0.1

0.2
0 Y
0
0.2
X −0.2

Fig. 1 (A and B) Two views of the pluteus larva of the sand dollar Dendraster excentricus at the 8-arm stage.
po ¼ postoral arm. al ¼ anterolateral arm. Ventral view (A) shows high arm elevation angles. Anterior view (B) shows
approximately biradial symmetry of arms. (From Grünbaum and Strathmann 2003.) (C) Model larva. Parameters
specifying form include length, number, and diameter of arms; arm elevation angle (a); angle between arms (b); center
of buoyancy (small sphere); and center of gravity (large sphere). Lateral ciliary bands on each arm are modeled as a
tangential velocity perpendicular to the axis of the arm with velocity varying with the cosine of the angle off the plane
containing the band (high to low velocities represented here by dark to light shading). For details see Grünbaum (1998).
(From Grünbaum and Strathmann 2003.)
314
Observed swimming in shear
Observations of 8-armed sand dollar larvae (Fig. 1A
and B) demonstrated that shear does affect swimming
of some armed larvae. In a chamber with vertical shear,
the plutei moved across flow lines from the upward
current to the downward current (Figs. 2 and 3). The
passively stable, upward-swimming larvae moved
R. R. Strathmann and D. Grünbaum
across flow lines toward downwelling water because
they tilted in the direction of least upward velocity.
The gradient of velocities was nearly linear across
most of the chamber (Fig. 2), and the horizontal velo-
city of the larvae was approximately constant as they
crossed the chamber (Fig. 4). The vertical velocity
of the larvae (combined swimming and water flow)
changed as they moved horizontally from upward to
downward flow. The plutei moved from the upward
to the downward current even though they were able
to reverse ciliary currents and thus potentially able to
maneuver.
The plutei also swam across flow lines from
upwelling to downwelling water when they were
anesthetized with 1:1 isosmotic MgCl2 and seawater.
Anesthetized larvae swim continuously with no
reversals and thus swim steadily (Strathmann 1971).
The most obvious difference from unanesthetized
larvae resulted from the narrow gradient of velocities
near the chamber wall. Drag from the chamber wall
produces a narrow velocity gradient, with zero velocity
at the wall. On the upstream side of the chamber the
anesthetized larvae moved toward the low velocity at
the chamber wall. The anesthetized larvae then became
“stuck” with arm tips against the wall, held against the
wall by their continuous swimming. This did not hap-
pen to non-anesthetized larvae, presumably because
they reversed ciliary beat and reoriented when they
contacted the wall. Anesthetized larvae did not become
“stuck” against the wall on the downstream side where
maximum downward velocities were less at the wall
than a small distance away. That anesthetized larvae
swam against the wall on the upstream side but not the
downstream side of the chamber is a further demon-
stration of movement across flow lines toward lower
Fig. 2 Eight-arm plutei of Dendraster excentricus in
vertical shear. Numbers at upper left are time in
seconds at each video frame. Water moves down at the
left (cooled side) and up at the right (warmed side).
A downward arrowhead marks a particle moving
downward and an upward arrow marks a particle
moving upward at 0, 4, and 8 s. The plot at lower right
shows a nearly linear velocity gradient, with the x-axis
matching the horizontal position in the video frames.
The pluteus marked with a star enters the field of view
moving upward, tilts in shear, and moves across the
shear gradient into downwelling water. At 16 and 20 s,
another pluteus (marked with O) enters at the bottom
of the field of view and moves toward the faster
downwelling current, whereas the third pluteus
(marked with X) is near the slower velocity near the
chamber wall and therefore moves upward near the
wall. The white scale line at the bottom of the 0 s frame
is 1 mm. The scale is not corrected for parallax for
objects through the depth of focus.
Good eaters, poor swimmers
Fig. 3 Trajectories of particles (small polygons) and the
8-arm pluteus larvae (large polygons) in vertical shear
during 40.12 s. The distance between spheres in tracks
indicates velocity, with faster upward flow on the right
and faster downward flow on the left. The interval
between polygons is 0.1 s. While swimming upward,
plutei were tilted by the shear and crossed flow lines
from the upward toward the downward flow. The scale
is not corrected for parallax for objects through the
depth of focus.
upward velocities: toward the wall on the upstream side
but away from the wall on the downstream side.
For the sample of all plutei traversing the line of
minimum particle motion at the middle of the cham-
ber, 87% of the plutei moved from the upward to the
downward current (Table 1). The plutei that did not
move across the chamber were those in the fast currents
near the chamber walls. Thus passive stability and
upward swimming can result in downward transport
in vertical shear (horizontal gradients in vertical flow
velocities). The mechanism is simple: larvae are tilted
toward the slower upward flow or faster downward
flow, and their swimming carries them across flow lines.
Modeled swimming in shear
Larval features were modeled to examine effect of dif-
ferences in larval form on performance in swimming
(Grünbaum and Strathmann 2003). The model
included number, length, diameter, and orientation
of thin, straight, rigid arms. The arms produced a
tangential current perpendicular to the axis of the
315
arm (see Grünbaum 1995). The model also included
a center of buoyancy anterior to a posterior weight
(Fig. 1C). Criteria for performance were upward velo-
city with no weight and weight capacity (the weight
that resulted in zero velocity) in still water and stability
in horizontal and vertical shear. These criteria were
chosen because many larvae swim upward or migrate
vertically (for example, Pennington and Emlet 1990;
Young 1995; Fenchel and Ockelmann 2002) and many
carry a heavy skeleton, shell, or rudiment of juvenile
structures. Speed and stability both affect a larva’s
ability to change its position in the water column.
Shear can reorient swimming larvae, changing the
direction of swimming away from the vertical or
other preferred orientation.
For still water, the results were presented as relative
speed or weight capacity. For larvae with the same total
arm length, the number of arms had some effect on
speed and weight capacity (Fig. 5). Speed and weight
capacity decreased as the elevation of arms above the
horizontal increased (Fig. 5). Performance in speed and
weight capacity was similar over a large range of low
arm elevations and declined most markedly at high
arm elevations.
The results in Figure 5 therefore indicate that low
arm elevations and few arms provide the greatest speed
and weight capacity in still water. Some ophioplutei
meet these criteria, especially at late stages, but many
ophioplutei and nearly all echinoplutei do not. In a
sample of sea urchin plutei from Mortensen’s (1921)
plates, 12 of 14 species from 4 orders had postoral
arms with elevations greater than 60 , despite a marked
decline of speed and weight capacity at arm elevations
above 60 (Fig. 5). Transverse portions of the ciliary
band between the arms may partly compensate for the
disadvantages of these high arm elevations, but why are
arm elevations high in so many plutei? One hypothesis
for the high arm elevations is problems associated with
stability in shear (Grünbaum and Strathmann 2003).
Horizontal shear (vertical gradients in horizontal
flow velocities) decreased upward velocity by tilting
larvae. With very strong horizontal shear, larvae
tumbled. The effect of horizontal shear was least for
larvae with many arms at low or high but not inter-
mediate arm elevation angles or with 2 arms at low
elevation angles.
Vertical shear (horizontal gradients in vertical flow
velocities) resulted in movement across flow lines, just
as in the experiment with real larvae. Larvae with high
arm elevations were less susceptible to this effect
of vertical shear. These results were for larvae with
symmetrically distributed arms. Some other arm
distributions avoided this effect of vertical shear
(D. Grünbaum, unpublished data). However, many
316 R. R. Strathmann and D. Grünbaum

Fig. 4 Horizontal and vertical velocities of the 8-arm pluteus larvae in vertical shear. Larval horizontal velocities changed
little as larvae crossed the nearly linear velocity gradient from upward toward downward current. Their vertical
velocities changed as they moved across the chamber from the upward current (right) toward the downward
current (left). The scale and velocities are not corrected for parallax for objects through the depth of focus.

Table 1 Frequency of direction of horizontal movement plutei have a nearly biradial or radial distribution
of 8-arm plutei of Dendraster excentricus traversing a of arms, like the sand dollar in Figure 1, and these
chamber with vertical shear
larvae did tilt and move horizontally from upward
Up to down Down to up to downward currents.
Seawater 39 5
1:1 Seawater and MgCl2 24 4 Extension to turbulence
Seawater 14 2
The swimming of model and real larvae in vertical
Horizontal direction was noted near the midline of the shear suggests that turbulence could decrease upward
chamber, where particle motion was minimal. Up to down is
movement from the upward to the downward current.
swimming. Though shear in turbulence is in many
Down to up is movement from the downward to the directions, the vertical component of shear would
upward current. move the passively stable, upward-swimming larvae
Good eaters, poor swimmers 317

Swimming speed
at neutral weight
0.4

2 arms
0.2 4 arms
6 arms
8 arms
10 arms
12 arms
0
Weight capacity 0 30 60

0.4
at zero velocity

0.2

0
0 30 60
Arm elevation angle in degrees

Fig. 5 Effect of arm elevation angle on speed (at zero weight) and weight capacity (at zero speed) in still water. Total arm
length and band length is the same for each larval geometry. Arm elevation angle is angle above the horizontal plane.
The model larvae are radially symmetrical and swim vertically. (From Grünbaum and Strathmann 2003).

toward the downward currents, giving this component Arm elevation angles in degrees
of turbulent motion a greater effect. The model indic- 11.3 33.8 67.5
0.5
ates that the effect of vertical shear is most pronounced
for larvae with low arm elevations (Fig. 6). High arm
0 2 arms
elevation angles provided greater stability in vertical
shear even though they performed less well in regard
after initial transients have passed)

−0.5
to speed and weight capacity. Turbulence may contrib-
ute to selection for high angles of elevation of pluteus
Average horizontal velocity

0 4 arms
arms, at the expense of lower performance in speed and
(for a fixed time interval

weight capacity.
−0.5
Predictions compared to larval forms
0 8 arms
The model provides an explanation of several features
of plutei. Two arms at low elevation angle provide high
−0.5
speed and weight capacity but low stability in vertical
shear (Figs 5 and 6). Some ophioplutei approximate
0 12 arms
this form, with 1 pair of arms much longer than the
others and at a low angle of elevation. In a sample of
−0.5
brittle star species from Hendler’s (1991) review, the 0 5 0 5 0 5 10
plutei with low elevation of the posterolateral arms Vertical shear
are those that retain these long arms as the heavy brittle (the change in vertical velocity along the x-axis)

star juvenile develops (Grünbaum and Strathmann
2003). The ophioplutei that approximate 2 arms at
low elevation appear to be achieving high speed or
weight capacity at the expense of stability in vertical
shear.
Other ophioplutei and nearly all echinoplutei have
high arm elevations. Some plutei compensate for
the high arm elevation with horizontal transverse
bands between the arms as lobes or epaulettes; some
Fig. 6 Plots from simulated trajectories in vertical shear.
Each horizontal axis is relative intensity of vertical shear.
Each vertical axis is the average horizontal velocity over
a fixed time interval after initial transients have passed
(t ¼ 100 to t ¼ 120). Negative values indicate
movement of larvae toward downwelling water. Solid
lines represent median velocity of 128 larval trajectories
in each plot. Shaded areas represent the 10th and 90th
percentiles of velocity. (From Grünbaum and
Strathmann 2003.)
318
ophioplutei rearrange the band into transverse bands
when they become competent to settle and develop
rudiments of juvenile structures (Grünbaum and
Strathmann 2003).
In most phyla, long arms or lobes of larvae have low
elevation angles. Examples include feeding larvae of
phoronids, brachiopods, and some gastropods. If the
model is an adequate approximation for these larvae,
then these larvae achieve greater speed and weight
capacity at the expense of stability in vertical shear.
Form is not the only feature affecting stability.
Stability can be increased by increased posterior weight
(Pennington and Strathmann 1990; Mogami and
others 2001), but greater weight increases stability at
the expense of speed in upward swimming.
The model predicts that for radially symmetrical
arm arrangements, there is no way to simultaneously
optimize speed (or weight capacity) and stability in
shear. Although real plutei depart from radial sym-
metry and can reverse ciliary currents, their forms and
their behavior in shear are consistent with this predic-
tion. The hypotheses presented by the model need fur-
ther testing. There are some arrangements of arms, less
symmetric, that avoid the transit from upwelling
toward downwelling currents (D. Grünbaum, unpub-
lished data). Also, the magnitude of the bias in larval
motion from turbulence, as opposed to steady shear,
needs to be investigated. Nevertheless, on present
evidence, it appears that long arms and lobes increase
maximum clearance rate with little investment in
ephemeral larval tissues (McEdward 1984, 1986;
Hart 1991, 1996), while decreasing stability in shear
(Grünbaum and Strathmann 2003). These results
help explain the loss of long arms in the transition
from feeding to non-feeding larval development in
many clades.
Shear and encounters
The model of armed larvae also explained how passive
stability results in crossing flow lines in vertical shear.
Models in which the organism’s swimming is added to
the water motion, as in models of effects of turbulence
on encounters (for example, Rothschild and Osborn
1988) have not included this consequence of passive
stability.
Moving water versus capturing
particles
A design criterion for a filter is that water pass through
it. A design criterion for a paddle is that water not pass
through it. Some arrays of cilia and setae both propel
the animal and filter water for food. For some of these
animals, there could be trade-offs between swimming
and feeding.
R. R. Strathmann and D. Grünbaum
Modeling by Cheer and Koehl (1987), Emlet (1988),
and Hansen and Tiselius (1992) demonstrates that
leakage of water through rows of cilia or setae increases
with a greater velocity of movement of the filtering
array through the water. Thus moving cilia or setae
faster is one way to improve performance as a filter,
with water moved per stroke decreasing. Another way
to increase leakiness and thus improve performance
in filtering is to tether the swimmer. Some flagellates
and some pelagic tintinnid ciliates may accomplish this
by attaching themselves to larger suspended objects
(Christensen-Dalsgaard and Fenchel 2003; Jonsson
and others 2004), but attachment is not the only
option. Gravity and drag can serve as tethers for
pelagic suspension feeders by restraining forward
motion (Emlet and Strathmann 1985; Fenchel and
Ockelmann 2002). Tethering results in a trade-off
between feeding and swimming. When the same
structure is being used for filtering and propulsion,
tethering enhances filtering while reducing speed.
A third possible means of increasing filtering may
be to oppose the filtering ciliary band with another.
Veligers of molluscs, many of the annelid trocho-
phores, and entoproct larvae capture particles between
parallel ciliary bands that beat toward each other
(Fig. 7). The effective strokes of cilia of the prototroch
(preoral band) are from anterior to posterior and
Fig. 7 Opposed prototrochal and metatrochal ciliary
bands of a larva of Sabellaria cementarium (Annelida,
Sabellariidae) in optical section. Black arrows show
direction of motion of effective strokes of cilia. White
arrows show direction of motion of curved recovery
strokes. Pr ¼ prototroch, M ¼ metatroch, fg ¼ food
groove.
Good eaters, poor swimmers 319

produce a current for both feeding and swimming. The feeders that are tethered to the earth can also increase
cilia of the metatroch (postoral band) are shorter than their capacity to clear particles by orienting to currents
the prototrochal cilia, and they beat from posterior to (for example, LaBarbera 1977). Though some annelid
anterior, toward the prototroch. Although the effective larvae associate with marine snow (Shanks and
strokes of the prototroch and metatroch thus propel Edmondson 1990), we know of no reports of larvae
water toward each other, the prototroch produces a with opposed-band feeding tethering to marine snow,
current for feeding and swimming because the com- other detritus, or the seabed.
pound cilia of the prototroch are much larger than
those of the metatroch. The prototrochal cilia move Drag as a tether
faster than particles within the length of their effective
Many veligers of bivalve and gastropod molluscs trail
strokes; particles beyond the reach of the prototrochal
mucous strands (Hamner and others 1975; Dobberteen
cilia are not captured (Strathmann and Leise 1979;
and Pechenik 1987; Hansen 1991; Fenchel and
Gallager 1988; Riisgård and others 2000). The inference
Ockelmann 2002). The strands reduce swimming
from these observations is that capture depends on the
speed and also change the flow field around larvae
prototrochal cilia overtaking particles.
(Fenchel and Ockelmann 2002). These larvae feed
Hypothesized mechanisms for the capture and
using the opposed beat of their prototrochal and meta-
concentration of particles include direct interception
trochal cilia. Presumably, shear around the proto-
by a cilium and sieving by more than 1 cilium in the
trochal cilia is increased by this mucous sea anchor,
effective stroke (Strathmann and Leise 1979; Gallager
enhancing filtration. Fenchel and Ockelmann (2002)
1988; Riisgård and others 2000). These suggested
compare the volume flows through the prototrochal
mechanisms are indirect inferences. The flow of water
ciliary band. From velocities of particles upstream from
has not been distinguished from the paths of particles
the prototroch, they calculate that the mucous strand
and cilia, and other possible mechanisms need to be
increases the volume flow through the ciliated band.
explored. For example, the paths of particles captured
In contrast, Emlet (1990) calculated that tethering
by sessile animals with opposed bands have suggested
decreased the overall flux of water past the prototroch,
other hypotheses, models for capture that depend
though it increased movement of the cilia relative to
on the oscillating flow from effective and recovery
the water and thus the volume swept by cilia. We have
strokes (Mayer 2001). Mechanisms of ciliary filtration
no explanation for the implied discrepancy in motion
remain speculative, but larval structures and behaviors
of cilia relative to the water but have followed Emlet’s
suggest trade-offs between filtration and swimming.
observations and analysis in our interpretation of
Here we focus on opposed-band feeding, adopt the
tethering.
hypothesis that clearance rates do indeed depend on
Other larvae trail mucous strands but for other
shear around the prototrochal cilia, and examine the
reasons than increasing shear around cilia. The feeding
implied trade-offs between feeding and swimming for
annelid larvae that trail mucous strands include chaet-
these larvae.
opterids and pisionids, which are reported to use the
Emlet (1990) has demonstrated that tethering does
strand itself to catch particles (Werner 1953; Åkesson
affect relative speeds of prototrochal cilia and particles
1961). Larvae of phyllodocid annelids trail mucous
for several larval forms, including one with the
strands (Fenchel and Ockelmann 2002), although
opposed-band feeding mechanism. Emlet tethered veli-
phyllodocid larvae do not feed with opposed bands
gers of the bivalve Crassostrea gigas on a microscope
(Strathmann 1987) and may capture particles by
slide and then simulated removal of the tether by sup-
other means than filtration. Some larvae trail strands
plying a flow equal to the swimming speed. Shear of
that aid in settlement (Abelson and others 1994).
water around prototrochal cilia was greater for tethered
larvae in still water. We examine several ways that lar-
vae may achieve this effect, with a trade-off between Weight as a tether
swimming and feeding. Although a primary function of calcified shells of veli-
gers appears to be defense against predators (Hickman
2001), the weight of the shells may also increase
Tether to an object clearance rates at the expense of a slower speed of
A connection to a large object can completely halt upward swimming. Gravity can serve as a tether
forward swimming and may further increase move- (Emlet and Strathmann 1985). An advantage of
ment of cilia or setae relative to the water because of mucous strands over shells is that the strands can be
drag from the nearby wall. For benthic suspension shed when speed is needed, but if a larva has a heavy
feeders, the large object is the earth. Ciliary suspension shell, there may be less need for a mucous strand.
320
An opposed band
The metatroch may be a built-in means of opposing
swimming to enhance filtration. The metatroch coun-
ters the push of water by the prototroch. It produces a
current toward the base of the prototrochal cilia in at
least some animals with the opposed-band system
(Strathmann and others 1972). If the metatrochal
cilia steepen shear gradients around the prototrochal
cilia in their effective strokes, then metatrochal beat
should increase filtration by the prototrochal cilia
(Strathmann 1987). The metatroch does not prevent
forward swimming, but its beat may reduce swimming
speed.
Some veligers hover when feeding, which suggests
that slowing swimming is associated with enhanced
filtration for these larvae (Fenchel and Ockelmann
2002). This seems counter to the prediction that arrays
of cilia become leakier when the effective strokes are
faster (Cheer and Koehl 1987; Emlet 1988; Hansen and
Tiselius 1992). One hypothesis is that swimming is
slowed by increased beat by the metatroch or changed
position of metatroch relative to prototroch during
feeding, but observations to test this hypothesis are
lacking. Increasing metatrochal beat or producing a
mucus strand in response to cues of food availability
might reduce swimming speeds when larvae are in
good foraging areas relative to their swimming speeds
in poor foraging areas. If this occurs, it would be
suggestive of an area-restricted search, a foraging
strategy known to promote aggregation in resource
concentrations in a wide diversity of organisms.
Telotrochs are locomotory bands of cilia near the
posterior ends of some larvae, including larval sabel-
lariid worms that feed with opposed bands. Telotrochs
and other transverse locomotory bands of cilia propel
the larva forward presumably reducing shear around
prototrochal cilia. Such specializations for swimming
appear to act in the opposite way to metatrochs and
may therefore represent compromise between swim-
ming and filtration but in favor of swimming.
Different phyla, different
swimming-feeding trade-offs
Different ciliary feeding mechanisms may result in dif-
ferent swimming-feeding trade-offs. Larvae of echino-
derms, hemichordates, phoronids, and brachiopods
produce feeding currents with bands of simple cilia,
usually about 20–25 mm in length. With simple cilia,
the flux across a unit length of band is limited; increas-
ing the maximum clearance rate requires a longer
ciliary band (Strathmann and others 1972; Emlet
and Strathmann 1994). These larvae thereby become
subject to swimming-feeding trade-offs involving arms
R. R. Strathmann and D. Grünbaum
or tentacles. However, these larvae capture particles
on the upstream side of the ciliary band that produces
the feeding-swimming current (Hart and Strathmann
1995) and therefore are not subject to the filter versus
paddle trade-off.
In contrast, nearly all larvae with opposed-band
feeding have compound prototrochal cilia. Grouping
cilia into compound cilia permits greater lengths and
greater tip velocities in effective strokes. Prototrochal
cilia of opposed-band feeders can therefore produce
a greater flux of water through a unit length of
ciliary band, and maximum clearance rates increase
with length of cilia as well as with lengths of ciliary
bands (Strathmann and others 1972; Strathmann and
Leise 1979; Gallager 1988; Hansen and Ockelmann
1991; Emlet and Strathmann 1994; Henderson and
Strathmann 2000). Fewer of the larvae with opposed-
band feeding have long lobes. They appear to be less
disposed toward instability in shear; however, their
filtration is downstream from the cilia producing
the feeding current, and these larvae are subject to
paddle-filter trade-offs.
Evolutionary consequences for
larval form
Instability in shear has implications for the evolution of
shapes of feeding and non-feeding larvae. Extension of
ciliary bands on arms provides a greater maximum
clearance rate. When the requirement for feeding is
lost, larval arms are usually reduced or lost entirely
(Strathmann 1974, 1978; Emlet 1995; Wray 1996;
McEdward and Miner 2001). Pelagic larvae, even if
non-feeding, have functional requirements for swim-
ming. Non-feeding larvae, like the feeding larvae, must
avoid predators and settle in favorable habitats, but
their ciliated lobes, if present, are short (as in non-
feeding larvae of gastropods) or mere vestiges of
the arms in ancestral feeding larval forms (as in
some echinoids). Arms could be lost because of drift
in the absence of selection (Strathmann 1975) or from
selection for enhanced performance in swimming
by means of transverse rings or broad fields (Emlet
1994). Instability in shear is another possible reason
that arms or lobes are commonly reduced or absent
in non-feeding larval forms (Grünbaum and
Strathmann 2003).
In contrast to larval arms or lobes, modifications
that enhance feeding with opposed ciliary bands at
expense of forward swimming are not anatomically
conspicuous. Many non-feeding spiralian larvae swim
with a prototroch and lack a metatroch with an
opposed beat, but determining if a metatroch is absent
requires close observation. Also, the metatroch may
Good eaters, poor swimmers
have other functions than capture of food, as suggested
by the retention of a reduced metatroch and food
groove in larvae of sabellid annelids, which lack an
open mouth to ingest food (Pernet 2003). Some larvae
with opposed-band feeding can modify the trade-off
through behavior, for example, producing or casting
off mucous strands or perhaps arresting the meta-
trochal but not the prototrochal beat (Strathmann
and others 1972). Thus there is often little difference
in form between larvae feeding with opposed proto-
troch and metatroch and related non-feeding pelagic
larvae.
Acknowledgments
We are grateful to the Friday Harbor Laboratories and
to NSF grants OCE 0217304 (RRS), and OCE 0220284
(DG) for support for this study and to T. Clay,
M. F. Strathmann, D. Vaughn, 2 anonymous reviewers,
and others for advice. This paper is dedicated to
Larry McEdward, the larval marvel.
References
Abelson A, Weihs D, Loya Y. 1994. Hydrodynamic impediments
to settlement of marine propagules and adhesive filament
solutions. Limnol Oceanogr 39:164–9.
Åkesson B. 1961. On the histological differentiation of the
larvae of Pisione remota (Pisionidae, Polychaeta). Acta Zool
42:177–225.
Cheer AYL, Koehl MAR. 1987. Paddles and rakes: fluid flow
through bristled appendages of small organisms. J Theor Biol
129:17–39.
Christensen-Dalsgaard KK, Fenchel T. 2003. Increased filtration
efficiency of attached compared to free-swimming flagellates.
Aquat Microb Ecol 33:77–86.
Dobberteen RA, Pechenik JA. 1987. Comparison of larval bio-
energetics of two marine gastropods with widely differing
length of planktonic life, Thais haemastomum canaliculata
(Gray) and Crepidula fornicata (L.). J Exp Mar Biol Ecol
109:173–91.
Emlet RB. 1988. Mechanical models of the opposed-band ciliary
feeding apparatus of invertebrate larvae. Am Zool 28:A33.
Emlet RB. 1990. Flow fields around ciliated larvae: effects of
natural and artificial tethers. Mar Ecol Prog Ser 63:211–25.
Emlet RB. 1994. Body form and patterns of ciliation in non-
feeding larvae of echinoderms: functional solutions to
swimming in the plankton? Am Zool 34:570–85.
Emlet RB. 1995. Developmental mode and species geographic
range in regular sea urchins (Echinodermata: Echinoidea).
Evolution 49:476–89.
Emlet RB, Strathmann RR. 1985. Gravity, drag, and feeding
currents of small zooplankton. Science 228:1016–7.
Emlet RB, Strathmann RR. 1994. Functional consequences of
simple cilia in the mitraria of oweniids: an anomalous larva of
an anomalous polychaete and comparisons with other larvae.
321
In: Wilson WH, Stricker SA, Shinn GL, editors. Reproduction
and development of marine invertebrates. Baltimore: Johns
Hopkins University Press. p 235–57.
Fenchel T, Ockelmann KW. 2002. Larva on a string. Ophelia
56:171–8.
Gallager SM. 1988. Visual observations of particle manipulation
during feeding in larvae of a bivalve mollusk. Bull Mar Sci
43:344–65.
Grünbaum D. 1995. A model of feeding currents in encrusting
bryozoans shows interference between zooids within a colony.
J Theor Biol 174:409–25.
Grünbaum D, Strathmann RR. 2003. Form, performance, and
trade-offs in swimming and stability of armed larvae. J Mar
Res 61:659–91.
Hamner WM, Madin LP, Alldredge AL, Gilmer RW,
Hamner PP. 1975. Underwater observations of gelatinous
zooplankton: sampling problems, feeding biology, and
behavior. Limnol Oceanogr 20:907–17.
Hansen B. 1991. Feeding behaviour in larvae of the opistho-
branch Phililne aperta. II. Food size spectra and particle
selectivity in relation to larval behaviour and morphology
of the velar structures. Mar Biol 111:263–70.
Hansen B, Ockelmann KW. 1991. Feeding behaviour in larvae of
the opisthobranch Philine aperta. I. Growth and functional
response at different developmental stages. Mar Biol
111:255–61.
Hansen B, Tiselius P. 1992. Flow through the feeding structures
of suspension feeding zooplankton—a physical model
approach. J Plankton Res 14:821–34.
Hart MW. 1991. Particle captures and the method of suspension
feeding by echinoderm larvae. Biol Bull 180:12–27.
Hart MW. 1996. The functional and evolutionary significance
of variation in feeding larval forms of echinoderms. Invertebr
Biol 115:30–45.
Hart MW, Strathmann RR. 1995. Mechanisms and rates of
suspension feeding. In: McEdward L, editor. Ecology of
marine larvae. Boca Raton: CRC Press. p 193–221.
Henderson SY, Strathmann RR. 2000. Contrasting scaling of
ciliary filters in swimming larvae and sessile adults of fan
worms (Annelida: Polychaeta). Invertebr Biol 119:58–66.
Hendler G. 1991. Echinodermata: Ophiuroidea. In: Giese AC,
Pearse JS, Pearse VB, editors. Reproduction of marine
invertebrates, Vol. VI, Echinoderms and lophophorates.
Pacific Grove, CA: Boxwood Press. p 355–511.
Hickman CS. 2001. Evolution and development of gastropod
larval shell morphology: experimental evidence for mechan-
ical defense and repair. Evol Dev 3:18–23.
Jonsson PR, Johansson M, Pierce RW. 2004. Attachment to
suspended particles may improve foraging and reduce preda-
tion risk for tintinnid ciliates. Limnol Oceanogr 49:1907–14.
LaBarbera M. 1977. Brachiopod orientation to water movement.
1. Theory, laboratory behavior, and field orientations.
Paleobiology 3:270–87.
322
Mayer S. 2001. Modelling of ciliary downstream collecting in
suspension-feeding invertebrates. Math Meth Appl Sci
24:1409–27.
McEdward L. 1984. Morphometric and metabolic analysis of the
growth and form of an echinopluteus. J Exp Mar Biol Ecol
82:259–87.
McEdward L. 1986. Comparative morphometrics of echinoderm
larvae. II. Larval size, shape, growth, and the scaling of feeding
and metabolism in echinoplutei. J Exp Mar Biol Ecol
96:267–86.
McEdward LR, Miner BG. 2001. Larval and life-cycle patterns in
echinoderms. Can J Zool 79:1125–70.
Mogami Y, Ishii J, Baba SA. 2001. Theoretical and experimental
dissection of gravity-dependent mechanical orientation in
gravitactic microorganisms. Biol Bull 201:26–33.
Mortensen T. 1921. Studies of the development and larval forms
of Echinoderms. Copenhagen: G. E. C. Gad. p 261.
Pennington JT, Emlet RB. 1986. Ontogenic and diel vertical
migration of a planktonic echinoid larva, Dendraster
excentricus (Eschscholtz)—occurrence, causes, and probable
consequences. J Exp Mar Biol Ecol 104:69–95.
Pennington JT, Strathmann RR. 1990. Consequences of calcite
skeletons of planktonic echinoderm larvae for orientation,
swimming, and shape. Biol Bull 179:121–33.
Pernet B. 2003. Persistent ancestral feeding structures in non-
feeding annelid larvae. Biol Bull 205:295–307.
Riisgård HU, Nielsen C, Larsen PS. 2000. Downstream collecting
in ciliary suspension feeders: the catch-up principle. Mar Ecol
Prog Ser 207:33–51.
Rothschild BJ, Osborn TR. 1988. Small-scale turbulence and
planktonic contact rates. J Plank Ecol 10:465–74.
Shanks AL, Edmondson EW. 1990. The vertical flux of meta-
zoans (holoplankton, meiofauna, and larval invertebrates)
R. R. Strathmann and D. Grünbaum
due to their association with marine snow. Limnol
Oceanogr 35:455–63.
Strathmann RR. 1971. The behavior of planktotrophic echino-
derm larvae: mechanisms, regulation, and rates of suspension
feeding. J Exp Mar Biol Ecol 6:109–60.
Strathmann RR. 1973. Function of lateral cilia in suspension
feeding of lophophorates (Brachiopoda, Phoronida,
Ectoprocta). Mar Biol 23:129–36.
Strathmann RR. 1974. Introduction to function and adaptation
in echinoderm larvae. Thalassia Jugoslavica 10:321–39.
Strathmann RR. 1975. Larval feeding in echinoderms. Am Zool
15:717–30.
Strathmann RR. 1978. The evolution and loss of feeding larval
stages of marine invertebrates. Evolution 32:894–906.
Strathmann RR. 1987. Larval feeding. In: Giese AC, Pearse JS,
Pearse VB, editors. Reproduction of marine invertebrates.
Vol. 9. Palo Alto: Blackwell. p 465–550.
Strathmann RR, Leise E. 1979. On feeding mechanisms and
clearance rates of molluscan veligers. Biol Bull 157:524–35.
Strathmann RR, Jahn TL, Fonseca JRC. 1972. Suspension feed-
ing by marine invertebrate larvae: removal of particles from
suspension by ciliated bands in a trochophore, pluteus, and
bdelloid rotifer. Biol Bull 142:505–19.
Werner B. 1953. Beobachtungen über den Nahrungserwerb und
die Metamorphose der Metatrochophora von Chaetopterus
variopedatus Renier u. Claparède (Polychaeta sendentaria).
Helgoländer Wiss Meeresunters 4:225–38.
Wray GA. 1996. Parallel evolution of nonfeeding larvae in
echinoids. Syst Biol 45:308–22.
Young CM. 1995. Behavior and locomotion during the dispersal
phase of larval life. In: McEdward L, editor. Ecology of marine
invertebrate larvae. Boca Raton: CRC Press. p 249–77.