Effects of Sea Anemones On the Flow Forces They Encounter

Organisms depending on moving fluid to transport them from place to place (such as various planktonic organisms or wind-dispersed seeds) have morphological features that maximize drag, whereas organisms that locomote through a fluid (such as birds or fish) tend to have structures that minimize drag. Sessile organisms (such as sea anemones or kelp) risk being dislodged or broken by drag forces, yet they depend on the fluid moving over them to bring them food and essential substances, to carry away their wastes, and to disperse their gametes or young.

Sessile organisms should therefore exhibit various compromises between minimizing and maximizing the effects of flow. I have investigated the ways in which the sizes, shapes and distributions of sessile animals affect the flow forces and feeding currents they encounter. There have been a number of studies of the distribution of benthic organisms in different types of flow (e.g. Southward & Orton, 1954; Ballantine, 1961; Gunjarova, 1968; Lewis, 1968; Hynes, 1970; Riedl, 1971; Schwenke, 1971), and a few biologists have actually measured flow velocities in the field (e.g. Charters, Neushul & Barilotti, 1969; Davies & Barham, 1969; Rhoads & Young, 1970; Grigg, 1972; Riedl & Machan, 1972; Terekhove, 1973), but the flow forces and the ways in which organisms affect them have not been investigated.

On the basis of the principles outlined below, I expected that the flow forces encountered by sessile animals would be determined by their structure and behaviour as well as by their locations in the environment. I made a comparative study of two species of sea anemones exposed to different flow conditions, the intertidal Antho-pleura xanthogrammica (Brandt) (Fig. 1C) which are found in areas exposed to extreme wave action (Rigg & Miller, 1949; Hand, 1955 a; Ricketts & Calvin, 1968; Dayton, 1971), and the large subtidal individuals of Metridium senile fimbriatum (Verrill) (Fig. 1 A, B) which occur in areas exposed to tidal currents (Hand, 1955b; Ricketts & Calvin, 1968; Kozloff, 1973).

Although adequate for estimating flow forces, the above equations do not fully describe the complicated forces on bodies in waves. For example, they do not consider that the eddies formed behind a body move over it when flow direction is reversed. Nonetheless, these equations are very useful in that they describe the dependence of flow forces on a few basic measurable features of a body and the fluid moving around it.

C_D = drag coefficient

C_M = mass coefficient

D = diameter of a body normal to the flow direction

F_D = drag force

F₁ = inertial force

F(t) = total force on a body in a wave at time t

P_p = period parameter

Re = Reynolds’ number = (ρUL)/μ, where L is a linear dimension of a body

S = projected area of a body normal to the direction of flow

T = period of a wave

U = velocity

U(t) = horizontal component of velocity of water in a wave at time t

= horizontal component of acceleration of water in a wave at time t

V = volume of a body

μ= viscosity of a fluid

ρ = density of a fluid

Sessile organisms in waves are subjected to pressure changes as well as to horizontal flow forces. The height of the water column above an organism, and hence the pressure on that organism, changes as a wave passes overhead, although such pressure changes are small. Larger pressures are exerted on shores by breaking waves. Large wave pressures (‘shock pressures’) last only a few thousandths of a second, are rare and cannot be predicted because their amplitude depends upon the particular shape and velocity of the wave and configuration of the shore at the moment of impact. It has been observed, however, that the pressures exerted on a shore by breaking waves are lower when air is entrained in the water, when the shore is horizontal or of gentle slope, and when the shore is cushioned by a layer of water such as the backwash of the previous wave (Wiegel, 1964: Carstens, 1968). I have not found A. xanthogrammica intertidally on exposed vertical rock walls normal to the direction of wave propagation where shock pressures are most likely to occur. Rather, I have found a few of these anemones on vertical walls parallel to the direction of wave propagation and many more on the horizontal bottoms of channels; these A. xanthogrammica are in the low intertidal zone where they are covered by a cushion of water when waves are breaking directly over them. I therefore do no expect A. xanthogrammica to encounter large pressures due to breaking waves. Furthermore, an increase in pressure around an anemone filled with incompressible water would not subject it to a mechanical load unless it was in the pressure gradient at the edge of a shock pressure zone; such a pressure gradient would tend to push the anemone laterally. Because the probability of an A. xanthogrammica encountering such a gradient is low, I have restricted the present study to the horizontal drag and inertial forces that accompany each wave.

By SCUBA diving and working low tides, I surveyed a number of sites along the coast of Washington to determine the characteristic habitats of M. senile and A. xanthogrammica and to observe their posturing and feeding behaviour. I assessed the shapes and dimensions of both species of anemones in the field as described in Koehl (1977).

I measured water velocities in order to characterize quantitatively the flow regimes of localities where the two species occur as well as of their microhabitats within those localities. With a thermistor flowmeter (LaBarbera & Vogel, 1976) capable of 1 mm spatial resolution, I was able to measure water speeds between the columns and tentacles of anemones. I used an electromagnetic flowmeter (EPCO Water Current Meter, Model 6130) to measure current velocities at exposed coastal sites. Flow probes were held rigidly in place by metal scaffoldings attached to the rock substratum in positions which did not interfere with the currents being measured. Simultaneous signals from both channels of either of the flow meters were recorded on CrO₂ magnetic tape using a portable stereo tape recorder (Sony TC-1261) and later transcribed using a chart recorder (Gould Brush 220) or an X–Y recorder (Hewlett-Packard, Model 7004B). The maximum standard error (S.E.) of repeated measurements transcribed in this way of the same velocity in a calibration tank was ± 3 % of the velocity for the electromagnetic flowmeter and ± 8 % for the thermistor flowmeter.

I constructed a force plate (Fig. 2) to make flow force measurements on anemones. Changes in resistance of X and Y strain gauges of the force plate were simultaneously monitored using two DC universal bridges (Bean, Model 101) and two strain meter-amplifiers (Bean, Model 502). Simultaneous signals from the two amplifiers were recorded and transcribed as described above. The force plate was calibrated by hanging weights from the specimen attachment point; the maximum standard error of repeated measurements of the same force recorded and transcribed as described was ±5%. Anemones were allowed to attach to circles of plexiglas (⁠ in. thick) which were then affixed to the force plate. The force plate could be bolted to the rock substratum in the field as well as to the bottom of a flow tank (Vogel & LaBarbera, in preparation). Water velocities in the flow tank were measured by timing particles ; the maximum standard deviation of repeated measurements of the same velocity was ± 8%. The flow tank was filled with Instant Ocean (33‰,) and kept at 10 °C.

The contributions to drag of various aspects of the shape, texture and flexibility of the anemones, was tested by using the force plate to measure the drag on models of anemones in the field and in the flow tank. The models were made variously from clay, laster, fabric, sheet aluminium, and flexible plastic foam.

Along the Pacific coast of the United States, Metridium senile range from small, shallow-water animals with large tentacles, to large, deeper-water animals with lobed oral discs and numerous small tentacles (Hand, 1955b). found the small M. senile in mussel beds at exposed coastal sites and on docks in Puget Sound, and the lobed large M. senile in the greatest abundance on vertical rock walls at depths greater than 10 m in Puget Sound. The few large M. senile along the outer coast were depths such that they did not experience the oscillating flow due to waves. The present study is limited to the lobed large M. senile.

The mean body proportions of expanded M. senile are reported by Koehl (1977). M. senile have long, slim, tapering columns topped by large, fluffy lobed oral discs. The most variable aspect of the shape of expanded M. senile is the diameter of the oral disc. M. senile are tall (mean height above substratum = 38·0 cm, standard deviation. S.D. = 9·0, n = 28). M. senile encounter steady tidal currents of relatively low velocity (the maximum velocities attained during the tidal cycles were less than 0·2 m.s⁻¹) (Fig. 3B). The mean ratio of mainstream water velocity to velocity at the oral disc of M. senile was 1·02, to velocity at the upper column, 1·09, and to velocity at the lower column, 1·17; these ratios are not significantly different from each other, however (F_[2,42] = 1·15, 0·25 < P < 0·50). Thus, because M. senile are tall, they are subjected to essentially mainstream current velocities (Fig. 3C).

M. senile in flowing water are bent over (Figs. 1 A, B): this bending orients their oral discs normal to the water current. I have observed these anemones filtering zoo-plankton from the water passing through the meshwork of numerous small tentacles on their lobed oral discs. Gut content analyses confirm that M. senile are zooplankti-vorous (Koehl, in preparation).

Anthopleura xanthogrammica commonly carpet the bottoms of intertidal surge channels at rocky exposed coastal sites. At high tide they are subjected to the oscillating bottom flow as waves pass overhead : at low tide they are subjected to the shoreward surge and seaward backwash of breaking waves. I also found a few solitary A. xanthogrammica at coastal sites where such wave action was attenuated by rock reefs located seaward of the anemones. The variance of the heights of A. xanthogrammica in exposed channels ((S.D.)² = 0·81) is significantly lower (F_{[25, 71]} = 7·11, P < 0·002) than the variance of the heights of A. xanthogrammica in protected areas ((S.D.)² = 5·76). The oral discs of A. xanthogrammica in groups carpeting exposed channels are all at essentially the same height; even though wider individuals tend to be taller (Fig. 4), wider anemones in exposed channels tend not to stand up taller than their narrower neighbours.

To assess whether the difference in height between exposed and protected A. xanthogrammica might be due to a structural difference between the two populations or to a transient difference in posture, I measured and collected individuals from both protected and exposed sites, put them in a sea table where flow velocity was < 0·05 m.s⁻¹, and measured their heights on the subsequent 8 days. The mean heights of exposed (2·1 cm) and protected (5·3 cm) A. xanthogrammica before they were collected were significantly different (F_[1,13] = 20·63, P < 0·001), whereas after 8 days in the sea table they were not (F_{[1, 13]} = 0·59, P < 0·50). After 8 days in nearly still water, the anemones from protected sites had not changed significantly in height (F_[1,18] = 0·04, P > 0·75), whereas the anemones from exposed channels were significantly taller (F_{[1, 18]} = 7·78, P < 0·025). This increase in height from a mean of 2·1 cm to a mean of 4·3 cm over a short period of time suggests that the difference in height between A. xanthogrammica from protected and exposed sites is due to posturing.

Current velocities over a shore are greatest during surge and backwash (Carstens, 1968). Mainstream flow in exposed surge channels during surge and backwash (Fig. 5 A, B) is bidirectional with a mean period of 7·7 s (S.D. = 0·9, n = 37). Because A. xanthogrammica in exposed surge channels are so short, flow over these anemones is of lower velocity than mainstream (Figs. 5 C, D). Flow over the anemones is also more turbulent than mainstream. The mean of the maximum velocities of each wave surge 40 cm above the A. xanthogrammica in a surge channel (0·51 m.s⁻¹) is significantly (F_{[28, 111]} = 174·65, P < 0·001) faster than the mean of maximum velocities at their oral discs (0·21 m.s⁻¹). The variance of the maximum velocities 40 cm above the anemones ((S.D.)² = 0·04) is significantly greater (F_{[28, 111]} = 4·0, P < 0·002) than that of maximum velocities at their oral discs ((S.D.)² = 0·01), indicating that the flow microhabitat of the anemones is more constant than is the mainstream flow regime of the area where they occur. Mean maximum velocity of surge at different heights above an A. xanthogrammica in a surge channel and the mean velocity of tidal current above a M. senile on a vertical rock wall are plotted in Fig. 6. Unlike the tall M. senile, which encounter essentially mainstream flow, the short A. xanthogrammica ‘hide’ from the maximum flow velocities which characterize the areas where they occur.

In contrast, the taller A. xanthogrammica in protected areas stick out into mainstream currents. For example, the mean maximum velocity of surge 40 cm above an A. xanthogrammica at a protected site was 0·37 m.s⁻¹ (S.D. = 0·17, n = 43) and that at the oral disc of this anemone was also 0·37 m.s⁻¹ (S.D. = 0·17, n = 154); the variance of maximum velocities ((S.D.)² = 0·3) is the same for mainstream flow and for flow on the anemones. Mean maximum surge velocity 40 cm above the oral discs of A. xanthogrammica at an exposed site (0·51 m.s⁻¹) was significantly faster (F_{[1, 123]} = 15·10, P < 0·001) than that of a protected site (0·37 m.s⁻¹) measured within one hour on thè same day. However, the mean maximum surge velocity on the oral discs of the ‘protected’ A. xanthogrammica (0·37 m.s⁻¹) is significantly faster (F_{[1, 263]} = 81·75, P < 0·001) than that on ‘exposed’ anemones (0·21 m.s⁻¹) measured within that same hour. This illustrates that a knowledge of mainstream flow conditions does not tell us the flow regime encountered by a particular organism.

A. xanthogrammica feed primarily on mussels which fall on their oral discs after being ripped off the substratum by starfish, logs, and waves (Dayton, 1973). A. xanthogrammica do not visibly deform in flowing waters; they remain upright in surge and catch the large prey items such as mussels which settle out of the slower water eddying over the anemones (Koehl, in preparation).

Although A. xanthogrammica in exposed channels are hidden from mainstream currents and M. senile in calm areas are not, A. xanthogrammica encounters flow velocities generally two to four times higher than those met by M. senile. I therefore fixnected the structure of A. xanthogrammica to minimize drag more than that of M. tenile The Re of a M. senile in a tidal current and the Re of an A. xanthogrammica in surge are both of the order of io⁴, hence pressure drag accounts for most of the drag on these anemones. Therefore, any shape, texture, or behaviour which minimizes wake size will minimize drag.

The flow forces on individual anemones were calculated using field measurements of their body dimensions and of flow velocities impinging on them. Each anemone was modelled as a combination of cylinders of the appropriate dimensions and the flow forces were computed as illustrated in the Appendix.

The calculations of flow forces on specific anemones (Koehl, 1976) and on ‘typical* anemones (Appendix I) indicate that although M. senile occur in calm regions and A. xanthogrammica in exposed areas, the flow force on an individual anemone of either species is nearly the same (about 1 N). I measured flow velocities over the anemones during the months of May, June, and July, 1975, and August, 1974, and no doubt velocities in surge channels during winter storms are higher than those I recorded. If the flow forces on exposed A. xanthogrammica are calculated (assuming they maintain the same shape during storms and assuming they encounter mainstream velocities) in a ‘worst case* situation (Riedl (1972) reports 5 m.s⁻¹ as the maximum velocity attained in coastal channels), the forces are of the order of 20 N. This is probably an overestimate because the anemones can change their shapes and sizes and because they most likely are not exposed to mainstream currents.

To test how realistic these rough predictions of flow forces on anemones were, the force plate was used to measure force on an expanded A. xanthogrammica in an exposed surge channel where velocity and body dimension measurements had been made. The mean maximum force of wave surge on the force plate alone was significantly lower (F[₁₍₈₀] = 45·57, P < 0·001, coefficient of determination = 0·36) than the mean maximum force on the plate plus the anemone. Hence the difference between mean maximum flow force on the plate plus anemone and on the plate alone, 1·2 N, is a reasonable estimate of the mean maximum force on the anemone. This estimate falls within the range of flow force values calculated for A. xanthogrammica.

I measured the drag on models of anemones in the flow tank at Re’s between io⁴ and 10^s to test the contributions of various aspects of shape, texture, and flexibility to drag. Drag was proportional to velocity² as would be expected for Re’s of 10⁴–10⁵.

A tall anemone should experience greater drag forces than a short one of the same diameter for several reasons. As mentioned above, the taller the anemone, the less it is hidden in the slower moving fluid near the substratum. The tall anemone also has a larger projected area normal to the flow than does the short one. Furthermore, the larger a cylinder’s ratio of diameter to length, the smaller the cylinder’s C_D, because fluid flowing around the free end of the cylinder reduces the size of the wake behind it (Rouse, 1961). The drag on a tall model of an A. xanthogrammica is greater than the drag on a shorter but otherwise identical model (Fig. 7). The ability of anemones to change their height thus enables them to modify the drag to which they are subjected.

An expanded anemone is not a plain cylinder, but has tentacles on its free end. It appears that tentacles make a large contribution to the drag on both A. xanthogrammica (Fig. 8) and M. senile (Fig. 9). Surge force measurements in the field for an A. xanthogrammica and for a model of the same dimensions, but without tentacles, indicated that mean maximum force on the anemone was 3-5 times greater than on the model.

Streamlined shapes minimize wake size, and thuspressure drag. A typical streamlined shape has a blunt nose and a long tapered tail. The rate of pressure increase along this tail is gradual and the point of boundary layer stall and separation occurs close to the rear tip of the body. The typical streamlined shape is effective in reducing drag only if the flow is from the blunt to the tapered end. It is therefore not surprising that organisms encountering unidirectional flow, such as fish and stream-bottom insect larvae, have the typical streamlined shape (Shapiro, 1961; Carstens, 1968; Hynes, 1970), whereas sessile organisms, such as anemones in tidal currents or waves where flow direction changes, do not. For a body in an area where flow direction changes, stream-lining (i.e. minimizing the rate of pressure increase along a body in flow) may be accomplished to some degree by any shape or orientation which presents most of the exposed surface area of the body parallel to the direction of flow (this also reduces the projected area of the body normal to the direction of flow). Considered in these terms, the shape of an A. xanthogrammica appears to be streamlined, whereas that of a M. senile bent over in flow with its oral disc normal to the direction of flow does no (Fig. 10). AS expected, most of the drag on models of expanded M. senile is due to drag on the oral disc (Fig. 11). Drag measurements on models also indicate that because water can flow between the lobes of a M. senile oral disc, drag is lower than it would be if the oral disc were not lobed (Fig. 11).

When the flow velocity over a M. senile in the field is artificially increased, the flexible oral discs of these anemones collapse like an inside-out umbrella, which probably reduces drag. Drag measurements on models of M. senile with rigid and with flexible oral discs illustrate this effect (Fig. n). Flexible M. senile in rapid flow are also bent down closer to the substratum where velocities, and thus drag forces, are lower. Thus, although the shape of a M. senile is far from being streamlined, the flexibility of these anemones offers a ‘safety mechanism’ whereby shape becomes more drag-minimizing as velocity increases, as has been shown for trees (Fraser, 1962). If the flow over a M. senile is artificially increased too much too abruptly, the anemone deflates and retracts into a hemisphere-shaped blob: this behaviour significantly reduces drag (compare Figs. 11 and 12).

Although a rough surface increases skin friction, it can sometimes reduce drag at high Re’s by moving the point of boundary layer separation rearward on a body. If a fluid is moving rapidly over a rough surface, turbulence may be created in the boundary layer, therefore momentum can be transferred to the slower layers of fluid which are less likely to stall and cause separation (Shapiro, 1961). The columns of A. xanthogrammica are covered with wart-like bumps (verrucae) which may affect drag. I found, however, that drag on a verruca-covered model of a retracted A. xanthogrammica was nearly the same as the drag on a smooth model of the same dimensions (Fig. 12).

Compliant surfaces and polymer coatings on solid bodies have been found to reduce drag in many instances, although the exact mechanism of this effect is still being investigated (Kramer, 1965; Wells, 1969; Mordvinov, 1972; Cox, Dunlop & North, 1974; Virk, 1976). The soft surface and the mucous coating of a sea anemone may reduce drag. The drag measured on a soft, slimy A. xanthogrammica, however, was not noticeably different from that on a hard clean model of the same shape and dimensions (Fig. 12).

The shape, size flexibility, texture, and behaviour of an organism affects the flow forces it encounters, as illustrated by the sea anemones M. senile and A. xanthogrammica. M. senile occur in calm areas, but because they are tall, they are exposed to mainstream current velocities. Although A. xanthogrammica occur in areas exposed to wave action, they are short and effectively hidden from mainstream velocities. Nonetheless, water velocities encountered by A. xanthogrammica are greater than those met by M. senile. However, the drag force on an individual of either species is about i N due to the respective shapes of each species.

I have used the case of these two species of sea anemones to illustrate the sorts of compromises made by sessile organisms between maximizing the transport done and. minimizing the mechanical forces caused by flow. M. senile are distributed in areas where flow velocities are low, but have an array of morphological features that maximize the effects of the currents they do encounter. In contrast, A. xanthogrammica are distributed in areas where flow velocities are high, but have morphological features that minimize the effects of flow. The water currents encountered by these anemones and their mechanical responses to the currents can be related to the manner in which the anemones harvest food from flowing water. M. senile bend over in currents and suspension-feed through their oral discs whereas A. xanthogrammica remain upright in surge and catch mussels which fall on their oral discs.

This work was supported by a Cocos Foundation Training Grant in Morphology, a Graduate Women in Science Grant, a Sigma Xi Grant, and a Theodore Roosevelt Memorial Fund of the American Museum of Natural History Grant. I wish to express my appreciation to S. A. Wainwright whose support and feedback made this work possible, and to S. Vogel who introduced me to fluid mechanics and who gave me much helpful advice. I am grateful to R. Fernald and A. O. D. Willows for the use of facilities at Friday Harbor Laboratories, University of Washington. I thank the U.S. Coast Guard and R. T. Paine for making it possible for me to work on Tatoosh Island, and M. Denny, M. LaBarbera, K. Sebens, and T. Suchanek for their invaluable help in the field. I offer special thanks to M. LaBarbera for all his help with instrumentation, and to N. Budnitz for his advice on statistical analyses.

Given : density (ρ) of sea water at 10 °C = 1026 kg.m⁻³ (Zerbe & Taylor, 1953); viscosity (μ) of sea water at 10 °C = 1·39 × 10⁻³ N.s.m⁻² (Sverdrup, Johnson & Fleming, 1942).

(Water temperatures in areas along the coast of Washington where the anemones are found range from 5 to 16 °C (Connell, 1970).)

I modelled a M. senile as diagrammed in Fig. 13 A, thus

I modelled a ‘protected* A. xanthogrammica as diagrammed in Fig. 13C.

The dimensions of this ‘typical’ protected A. xanthogrammica were calculated from the mean height of protected A. xanthogrammica. (⁠⁠, S.D. = 0·02, n = 25) and using the body proportions listed in Table 2 of Koehl (1977). The flow velocity used on the oral disc was the mean maximum velocity (⁠⁠, S.D. = 0 · 17, n = 154) of surge recorded at the oral disks of protected A. xanthogrammica during spring and summer months. The velocity on the column was calculated using the mean ratio of column velocity to oral disc velocity (⁠⁠, S.D. = 19, n = 5) for protected anemones. The mean maximum acceleration at the oral discs of protected A. xanthogrammica (⁠⁠, S.D. = 2 · 62, n = 118) was used in these calculations, and the acceleration on the column was estimated to be 2-98 m.s^-2. The period of wave surge in protected áreas (⁠⁠, S.D. = 1 · 8, n = 15) was used in calculating P_p.

The Re for maximum velocity flow in this A. xanthogrammica was calculated to be 3·5 × 10⁴, thus the flow forces were calculated in the same manner as they were for the ‘exposed’ A. xanthogrammica above. The calculated F_{max total} was 2·84 N.