ABSTRACT
The swimming behaviour of the anemone Stomphia coccinea, previously known to be a specific response to contact with two species of starfish, is briefly described.
The sphincter, isolated or intact, gives a reflex quick contraction when the tentacles are stimulated with whole Dermasterias imbricata or with extract, but not when other starfish are used. Sphincter preparations stimulated electrically give a quick facilitated contraction at frequencies above 1 in 3 sec., and slow contractions at fre quencies below 1 in 2 sec. Quick and slow contractions appear to have the same threshold, which is unaffected by the presence of starfish extract. Possible attributes of the chemosensory system are discussed.
Elongation of the column is an essential preliminary to swimming, whereas detachment from the substratum, due mainly to contraction of the parieto-basilar muscles, does not always occur. Cutting experiments, and orientated recording of the swimming movements which ensue, suggest that excitation is maintained locally in the column, and that the parieto-basilar and circular muscles function synergically. Multipolar nerve cells in the column may be concerned with this activity.
Various special features of the anatomy of Stomphia may be correlated with its ability to swim.
Although the origin and adaptive value of the swimming response are still obscure, it is clear that it must be of distinct biological advantage to the species.
INTRODUCTION
Sea anemones are a group of animals whose diagnostic structural features are remarkably constant. Their diverse behaviour patterns can nevertheless be related to anatomical differences, especially in the distribution of muscular system and nerve-net. The arrangement of the nervous system and sense cells is less readily detected than that of muscles, for these elements are diffuse and require special staining methods for recognition. As yet their structure is known in detail only in Metridium(Pantin, 1952 ; Batham, 1956; Batham, Pantin & Robson, 1960), and variations in other anemones corresponding to observed physiological differences would not be unexpected.
Properties of the neuromuscular and sensory systems can be analysed most readily in rapid responses to external stimuli, such as the reflex closure of the sphincter in Calliactis(Pantin, 1935 ; Passano & Pantin, 1955). Inherent slow activity, on the other hand, which has been demonstrated in Metridium(Batham & Pantin, 1950b, c) and in Calliactis(Needier & Ross, 1958) has yet to be interpreted in terms of sensory and nervous activity.
Responses vary unexpectedly from one species to another. The swimming behaviour of Stomphia coccinea is a response to contact with certain starfish and could not be predicted from what is known of other anemones. The unusual circumstances of this behaviour were described by Yentsch & Pierce (1955), and Sund (1958) and Wilson (1959) have contributed more recent studies. In the present paper this behaviour pattern is examined further, and an attempt made to interpret the sequence of activity in underlying structural components. It may later be possible to relate various special features more satisfactorily to what is known of the behaviour of other actinians.
MATERIAL AND METHODS
Stomphia coccinea belongs to the deep-water family Actinostolidae, and has been described by Carlgren (1893, 1921) and Stephenson (1935). It ranges from Siberia through Scandinavia to the Pacific, and is infratidal in relatively cold waters, having been recorded between 9 and 445 m. (Stephenson, 1935).
The present observations were carried out at the University of Washington Marine Laboratories at Friday Harbor. Anemones were usually obtained from President Channel at a depth of about 200 m., where the summer temperature of the water probably does not attain 12°C. and considerable bottom currents result from tidal changes Many specimens were found on shells of Modiolus modiolus, and they were kept in running sea water at 10− 15°C. and fed occasionally. Dermasterias imbricata, one of the starfish which causes the anemone to swim, is a Northern Pacific species which may be collected intertidally. It did not occur in dredge hauls containing Stomphia.
A kymograph and a Kodak Cine Special 11 were used for recording. Details of methods are given in relevant sections.
SEQUENCE OF BEHAVIOUR
The general pattern of swimming behaviour in Stomphia has been described by Yentsch & Pierce (1955), Sund (1958), Wilson (1959), and Hoyle (1960). It is summarized here and in Text-fig. 1 for reference in the subsequent analysis.
Stimulus
Surface of the starfish Dermasterias imbricata in brief contact with the tentacles (up to 10 sec.). Hippasteria spinosa, the other effective starfish (Sund, 1958), was not available.
(1) Initial response
Tentacles adhere to the starfish on contact. After several seconds they contract, followed at once by contraction of the sphincter and retractor muscles.
(2) Elongation
As the sphincter relaxes, a wave of elongation passes down the column. The tentacles and disk reappear and expand to their fullest extent.
(3) Detachment
At the same time the foot is released from the substratum, and unless it is very firmly attached it decreases in diameter. The centre of the pedal disk, initially convex, becomes delimited by a deepening circular groove, and is constricted to a conical projection.
(4) Swimming
A series of abrupt bending movements of the column ensues, which may cause the animal to ‘swim’. Although the frequency of the contractions falls off, they may continue for several minutes. The bending movements may be preceded by vigorous whirling of the column about its base.
(5) Recover
An inactive period follows, during which the anemone resumes its normal shape. During this time it may not respond to renewed stimulation. The surface of the pedal disk presently exhibits a marked facility of adhesion, and as it reattaches to the substratum the ordinary sessile position is regained.
ANATOMICAL STRUCTURE
The following account of the anatomy of Stomphia agrees with descriptions given by Carlgren (1893, 1921) and Stephenson (1935), and refers to parts of the animal which will be discussed later. Several features can be correlated with the swimming behaviour.
The general plan of the anemone is shown in Pl. 1 and Text-fig. 2. There are no cinclides, and the only opening to the coelenteron is the mouth. None of the anemones examined was found to have a pore in the centre of the pedal disk like that shown by one of Yentsch & Pierce’s specimens.
Below the tentacles is a strong mesogloeal sphincter muscle (Text-fig. 2, Pl. 2 a). Although continuous with the endodermal circular muscle of the column, which is also well developed, the sphincter is clearly defined and can often be distinguished in living specimens by the presence of a pink pigment in the muscle fibres. This could be an actinohaeme such as occurs in certain other deep-water anthozoans like Tealia(Fox & Pantin, 1944). The circular muscle continues into the foot, oral disk and tentacles.
As seen in Pl. 1d, there are usually eighteen pairs of perfect mesenteries. Six primary and twelve secondary pairs may be distinguished by the corresponding arrangement of tentacles. Incomplete mesenteries develop progressively all the features of complete mesenteries, and the youngest cycles show the unequal development of a pair which characterizes the Actinostolidae (see Pl. 1d, e). None of the mesenteries reaches the centre of the pedal disk, which is thus free (Pl. 1 a, e).
The perfect mesenteries carry three different muscles. The strong longitudinal retractors are endocoelic except in the directive mesenteries, and resemble those of other anemones (Text-fig. 2, Pl. 1b, 2b). On the opposite face there are transverse fibres of the radial muscle. In addition, there is on this surface a well-developed parieto-basilar muscle, which originates along the base of the mesentery. It is seen in Pl. 1 b as a triangular area inserted below the tentacles about two-thirds the way up the column. As the parieto-basilar muscle tapers towards the upper part of the column, the mesogloea of the septum between it and the sphincter is thickened into a substantial ridge which might almost be regarded as a tendon (Text-fig. 2, Pl. 1c). The parieto-basilar muscle develops as a fold which grows over the radial muscle fibres on the same side of the mesentery. The relative positions of these two muscles and of the retractor are shown in Pl. 2b.
In the pedal disk, endodermal basilar muscles run on each side of a mesentery at its junction with the foot. The radial muscle of the oral disk, on the other hand, and the longitudinal muscle of the tentacles, are ectodermal. As in certain other anemones, however, they are sunk into the mesogloea.
THE INITIAL RESPONSE
The first part of the response was studied in preparations of the sphincter with the crOwn of tentacles intact These were prepared from anemones anaesthetized for half an hour with equal parts of MgCl2.6H2O and sea water (Batham & Pantin, 1951), and left in running sea water for 24− 36 hr. before use. Experiments were carried out at temperatures between 9° and 12° C.
Nature of the stimulus
The specific quality of the sensory stimulus provided by starfish which release swimming behaviour has already been pointed out (Yentsch & Pierce, 1955; Sund, 1958). Sund made numerous laboratory and field tests which established that only two out of many local starfish, and none of the common invertebrates found in the same locality as Stomphia, would cause swimming. Preliminary work with Dermasterias imbricata and Hippasteria spinosa, both of which provoke a response, suggested that some parts of the starfish, in particular the aboral surface, were more effective than others. Further analysis of the stimulus provided by starfish is in progress (Ward, 1958, and curent work).
When the tentacles are stimulated by the aboral surface of Dermasterias, they adhere to the starfish, and after a latent period of at least 5 sec., both tentacles and sphincter contract Control tests such as rubbing the tentacles with a glass rod, and contact with various other starfish, cause no reaction. Records from an isolated sphincter preparation are shown in Text-fig. 3. The initial sphincter reflex thus indicates the specificity of the whole response. The stimulus cannot be a purely mechanical one, and must have a chemical component to which the tentacles are sensitive.
From work in progress (Mr J. Ward, which he kindly allows me to quote), it is known that an extract prepared by homogenizing the aboral surface of Dermasterias in sea water releases swimming in Stomphia as effectively as whole starfish. Such an extract applied near the tentacles of sphincter preparations causes a contraction like that produced by contact with starfish (Text-fig. 4). The extract was freshly prepared according to Ward’s method, by homogenizing 1 g. of aboral epidermis in 3 ml. sea water for 5 min. and centrifuging. A few drops of the clear supernatant would cause an anemone to swim vigorously. In testing sphincter preparations, 0 · 5 or 1 ml. were delivered about 1 cm. away from the tentacles with a syringe. After a test, the sea water was changed twice and the preparation allowed to recover for 30 min. This con-trms that the sphincter response can be evoked by a purely chemical stimulus.
The effectiveness of extract appears to fall off rapidly with dilution. In one experiment, 1 ml. of extract which had been diluted 1 in 10 had no effect. Control tests using only 0 · 5 ml. of undiluted extract (prepared as above) provoked sphincter contractions both before and after the trial of ineffective diluted extract. Such results indicate a sharp concentration threshold. When whole Dermasterias is used instead of extract, a high local concentration of the chemical substance secreted at the surface of the starfish presumably obtains on contact with the tentacles, and it is possible that the mechanical stimulus also provided by contact may be irrelevant. It is likely that these characteristics of the response, and its relatively short latency, reflect the properties of receptor elements present at least in the tentacles.
Properties of the sphincter muscle
A few observations were made on the response of sphincter preparations to electrical stimulation. Shocks were given with a Labtronics thyratron-discharge stimulator, using silver/silver chloride electrodes, and contractions were recorded with a light spring lever. The effects of electrical stimulation have been more extensively examined by Hoyle (1960).
Preparations of the sphincter muscle show both quick and slow contractions. As in Metridium(Batham & Pantin, 1954; Ross, 1960) and Calliactis(Pantin, 1935 b; Ross, 1957), quick contractions, showing facilitation, are obtained at frequencies above 1 in 3 sec., whereas slow contractions only are obtained at frequencies below this. A slow contraction may, however, follow a quick one, and at shock intervals of 1 or 2 sec. both are obtained. These effects are shown in Text-fig. 5. As in Metridium and Calliactis, quick and slow contractions appear to have the same electrical threshold (see Text-fig. 7 A).
A sphincter preparation which has been stimulated for some hours may showspon-taneous slow contractions. In addition, quick contractions soon begin to show a slow component as delayed relaxation. The example shown in Text-fig. 6C recalls similar observations by Batham & Pantin (1954) and Ross (1957) for Calliactis, by Batham & Pantin (1954) for Metridium, and by Horridge (1958) for Cerianthus. The relation between quick and slow types of contraction in such muscles, where only one type of muscle fibre is present (Grimstone, Home, Pantin & Robson, 1958), is not yet clear.
Nature of the sphincter response
Records of sphincter contractions caused by whole Dermasterias or by extract resemble quick responses to electrical stimulation, with the difference that relaxation is invariably rapid, whereas this is but rarely the case after electrical excitation. Text-fig. 6 illustrates two comparable contractions, and also a well-marked compound response. The characteristically rapid contraction and relaxation caused by Dermasterias must somehow be related to the pattern of innervation of the sphincter. The histology of this region is difficult, however, and relevant details are not yet available for any actinian. It may be noted that however carefully a mesogloeal sphincter muscle of the kind present in Stomphia or Calliactis is dissected out, some circular muscle is included in the preparation.
Dermasterias extract, whether concentrated and producing a contraction, or diluted 1 in 10 and failing to do so, appears to have little effect on the electrical threshold Text-fig. 7B shows a test of the quick response to two shocks 1 sec. apart; the small change in threshold would not be unusual in the absence of extract. The unchanged neuromuscular reaction of the sphincter in the presence of extract is further evidence that the specific response to Dermasterias may be attributed to receptor elements. The hct that immediately after a response to Dermasterias the anemone is inert to renewed stimulation with starfish or extract, suggests further that sensory adaptation to the starfish substance may occur. It should be pointed out, however, that Anemonia has been shown to adapt slowly to the chemical stimuli of food substances (Pantin & Pantin, 1943). Most vertebrate chemoreceptors are considered to show slow sensory adaptation, although in the mammalian tongue the initial frequency of impulses from single fibre preparations of chemoreceptors falls off rapidly to a maintained lower level (Pfaffman, 1941 ; Beidler, 1953). Conclusions cannot therefore be drawn with regard to Stomphia without further experiments.
Circumstantial evidence for the mode of action of Dermasterias chemical is provided by the response of independent effectors. The adhesion of tentacles, indicating the discharge of nematocysts, seems to parallel the sensory response. In these experiments, the tentacles did not adhere to glass rods, to ineffective starfish (except Henrida, which may occasionally cause swimming (Ward, 1958)), or to Dermasterias during the time after a response during which renewed stimulation had no effect. This suggests that the Dermasterias chemical may have a general effect at the tentacle surface rather than a particular one on receptors. Existing knowledge of how chemical and mechanical stimuli are concerned in the feeding response (Pantin & Pantin, 1943) and the discharge of nematocysts in Anemonia(Pantin, 1942) does not contradict this idea.
EXPANSION
After stimulation by Dermasterias the anemone expands much beyond its normal resting size. As the sphincter relaxes, a wave of contraction passes towards the foot, and the column elongates rapidly. This is accompanied by complete relaxation of the crown of the anemone, which becomes expanded to the point of translucence. This condition is not produced by any other ‘natural’ stimuli, such as feeding or the absence of light.
Elongation of the column is due to contraction of the circular muscle, whose tone is maintained throughout swimming (see p. 357). Circular contraction is propagated from the sphincter region after the initial response : this may be demonstrated in a midcolumn strip of circular muscle, which contracts slowly when a small tag of sphincter and tentacles at one end is stimulated.
The expansion of the disk, sphincter and tentacles seems to be due to hydrostatic pressure maintained by the contracted circular muscle of the column, for isolated sphincter preparations do not relax beyond the initial baseline after stimulation with Dermasterias, as they might if some kind of neuromuscular inhibition occurred. Once a responding anemone begins to elongate, water rapidly enters the coelenteron through the mouth ; one anemone, for example, took in water equal to one-sixth of its previous volume.
DETACHMENT
This phase is variable, as an anemone may show swimming movements without detaching from the substratum. More usually it will appear to detach itself with a vigorous upward jerk. In some cases, the reaction may spread from one edge of the pedal disk, or the foot may contract evenly in diameter as it lifts away from the surface. Sometimes the process is incomplete.
Observation of anemones detaching from a glass surface suggest that this activity has three components :
The ‘jerk’ is due to contraction of the parieto-basilar muscles, which lift the pedal disk from the surface. The centre of the foot, which has no mesenteries attached (p. 345), becomes convex owing to the increased coelenteric pressure and thrusts the anemone upwards. If the parieto-basal contraction spreads more slowly from one side, forming a concentric groove (see p. 344), detachment proceeds gradually.
At the same time, the pedal disk decreases in diameter. Although this is partly due to the wave of circular contraction which reaches it from the column, the rate at which this process occurs suggests that the basilar muscles may contract as well (Textfig-8; p. 346).
Slow motion films show that a wave of opacity passes very rapidly across the pedal disk before the anemone is released from the surface. This might be a local contraction of the epithelium (see below).
These features can be related to the structure of the pedal disk. As shown in Pl. 2 c, fibres of the parieto-basilar muscles pass right through the mesogloea of the disk. This condition is not usual in actinians and must increase the effectiveness with which the parieto-basilars release the anemone from a surface. In addition, the ectodermal epithelium contains numerous cells which give the impression of tough vertical strands. As seen in Pl. 2d, the proximal parts of many of the cells stain strongly with acid fuchsin. The basal parts of the cells end bluntly in the mesogloea, and the epithelial surface is covered by a cuticle. Batham & Pantin (1951) give a similar picture for Metridium. If the array of vertical strands shown in Pl. 2d is contractile, it may be responsible for the local wave of epithelial detachment noted above. There could, moreover, be functional co-ordination with the parieto-basilar system, since the fibres of these muscles appear to make contact with the epithelium.
SWIMMING
Orientation
Whether or not detachment occurs, actual swimming movements are preceded by elongation of the column. They may be recorded from an anemone held in a fixed position. Text-fig. 9 gives some idea of swimming activity shown in these conditions. The anemone is fastened to a wax dish a day or two before, by means of fine pins ; the dish is then held stationary in a large vessel of sea water by means of vaseline. Two threads, previously sewn into the edge of the disk at opposite points, are attached to light spring levers and record movements of the column.
Text-fig. 9A is part of a slow kymograph record of a swimming response after a Dermasterias was placed in contact with the tentacles of the anemone for about 10 sec. (Stim). The initial retraction shows as a small hump (sph.) at the beginning of the lower tracing. Re-expansion and elongation of the anemone are indicated by the rising baseline of the upper tracing in particular, and both levers then record the ensuing series of bending movements of the column. Swimming movements may continue for several minutes, during which, as may be seen in Text-fig. 9 A, their frequency falls away. Discrete contractions of the parieto-basilar muscles make the column bend first to one side, then to another, in rapid succession. Although Textfig. 9 A suggests that the contractions vary in amplitude, the apparently irregular height of successive contractions reflects instead their orientation in relation to the position of recording threads on the disk. A parieto-basilar contraction on the same radius as a recording thread gives a maximal deflexion of that lever, for example, but will hardly affect the tracing of the other lever; this is seen in the last contraction of the record in Text-fig. 9 A. On the other hand, a bending movement directed midway between the attachments of the recording threads will deflect both levers equally, as in the first two parieto-basilar contractions of this record, and so on. Text-fig. 9 A therefore shows that bending movements occur at a variety of points on the circumference in relation to the position of the attached threads.
A faster kymograph record of the swimming response, such as that shown in Textfig. 9B, shows that while the frequency of parieto-basilar contractions is irregular, a short interval between two bends (sometimes less than a second) is often followed by a longer one. This gives the impression of paired contractions followed by a pause. There are three examples of this in Text-fig. 9B. Two such contractions are in general at diametrically opposite points of the disk, whereas it may be said that two successive bends occurring less than 180 ° apart are not usually paired in time. The record in Text-fig. 9B may be interpreted as follows:
The tentacles of the anemone were stimulated by contact with the arm of a Dermasterias(Stim). After about 8 sec. the initial sphincter closure and retraction reflex ensued (sph.), followed by re-expansion and elongation of the anemone (note change in baseline level). The first parieto-basilar bending movement (pb.) was towards a point roughly equidistant from the two recording threads. After a pause, there were two paired contractions, and another single one. The last four contractions of the record fall into two pairs, and the differences between the upper and lower tracings indicate the orientation towards opposite radii of the members of a pair.
The orientation sequence of the parieto-basilar contractions has been examined further. If an anemone pinned to wax is stimulated (with Dermasterias) to swim, the directions of bending movements can be marked on a piece of paper beneath the disk with a recording error of not more than 5 °. It was found convenient to use stencilled diagrams marked with points of the compass at 60 ° intervals. As indicated by the inset diagram of Text-fig. 10, the anemone is centred on one of these with the directive axis as a reference line. For a complete record, the sequence of swimming movements is timed on a kymograph with a tapping key, while the direction of bending is marked of the blank stencil. From these data, the orientation of successive contractions can be plotted against time, as shown in Text-fig. 10. All the anemones were placed with their directive axes along the 0 − 180 ° axis of the stencilled compass grid. The direction of bending, i.e. the radius on which any parieto-basilar contraction takes place, is then measured as an angle in relation to this axis. In the graph shown in Text-fig. 10 each sharp peak indicates a point on the circumference of the anemone at which a parieto-basilar contraction took place, and the zig-zag line joining these points represents their chronological sequence. It will be seen that the angle between successive points (representing the sites of consecutive parieto-basilar contractions) is of the order of 180°; the beginning of the graph shows this fairly well. With the conventions used here, variation in the main part of the record ranges from about 90° to 270°.
In Text-fig. 10 the mid-points of lines joining successive peaks are shown by open circles. If the swimming anemone, bending first to one side and then to the other, maintained these parieto-basilar contractions at exactly the same opposite radii throughout, the open circles would form a horizontal line. As in fact most of the graph presents a steady slope, it can be seen that the sequence of approximately opposite contraction radii itself progresses slowly round the anemone at a speed rather greater than one revolution per minute. The time scale of this rotation is of the same order in all the records obtained. In Text-fig. 10 the sequence performs a total of almost five revolutions. The swimming movements continue for 6 min., during which (as already seen in the records of Text-fig. 9 A), their frequency declines.
The significance of this slow phenomenon is not clear. The slow migration of the site of parieto-basilar contractions may be established only after some initial oscillation, as in the graph of Text-fig. 10. It may also reverse its direction round the anemone at any stage during swimming. The precise speed and direction of rotation seem, moreover, to be independent both of the interval between contractions, and of their actual orientations. Whatever basic excitatory process in the neuromuscular system maintains the swimming contractions (see p. 358), it is difficult to attribute this slow shift in the orientation of parieto-basilar activity to a distinct, equally specific process. If, for example, steady ‘rotation’ represented by the slope of the graph in Text-fig. 10 were due to a slowly conducted wave of contraction progressing round the circular muscle of the column, some initial oscillation might be expected, but not the reversals from the clockwise to anticlockwise, or vice versa, that occur in several swimming records. The phenomenon is too slow and too variable to be explained as specific activity within the nervous system or in the muscular system. Despite the well-defined slope of graphs such as that shown in Text-fig. 10, it does not seem likely that any special excitatory process responsible for the appearance of rotation in the bending sequence exists. Some kind of asymmetry in the main excitatory process may provide an explanation.
It may also be noted that graphs of the kind shown in Text-fig. 10 cannot be interpreted as a repetitive pattern associated with particular mesenteries. The six primary anatomical radii of the anemone have no special significance in this context, although it may be recalled that there are, in fact, usually eighteen pairs of perfect mesenteries (p. 345), whose parieto-basilars are those mainly responsible for the swimming movements.
In a vigorous response the column of the anemone often performs a whirling move-ment about its own axis before the usual abrupt bending movements begin. This whirling is recorded by Yentsch & Pierce (1955), and by Sund (1958), and represents the serial contraction of parieto-basilar muscles round the circumference of the anemone. This activity, if rapid, is the one which initially propels the animal into some semblance of ‘swimming’; for the staccato contractions described above do not generally raise the anemone in the water to any extent.
Transmission of excitation
It has been observed by Wilson that both halves of a horizontally transected anemone may show the components of a swimming response to Dermasterias. This has been broadly confirmed, although preparations are no less variable than whole specimens. Vertical halves can also give a full response. One vertical half, and a horizontal half which included the lower pharynx and pedal disk, both continued to give proportionately accurate swimming responses for more than 5 weeks, by which time some regenerative growth had occurred. The latter specimen also demonstrated that there must be receptors sensitive to Dermasterias elsewhere than on the disk and tentacles (see p. 348 and Text-fig. 11A, B).
A variety of cutting experiments was carried out to determine which parts of the animal could propagate the sequence of parieto-basilar excitation. It was found that bending movements would still occur at opposite radii when these were joined only by mid-column tissue; as in Text-fig. 11 C, the disk,’pharynx, upper column, and foot may be cut, and yet stimulation of the tentacles on one side with Dermasterias still leads to parieto-basilar swimming contractions which may take place at any point on the circumference. If only the foot, oral disk, or sphincter, or a combination of these, were transected, the sequence did not appear to be affected either. When preparations gave responses of sufficient duration for the plotting of activity graphs of the kind shown in Text-fig. 10, the progress of excitation appeared to be fairly normal. These experiments, although far from conclusive, suggest that excitation of the parieto-basilar muscles during swimming is maintained in the column.
It is relevant that the excitation of circular and parieto-basilar muscles appears to be closely correlated. When the tone of the circular muscle is high, and the anemone is fully extended, swimming movements are very vigorous. As the frequency of bending falls off towards the end of swimming circular tone is lost and the anemone increases in diameter, ultimately regaining its resting dimensions. As in Metridium(Batham & Pantin, 1954; Robson, 1957; Batham, Pantin & Robson, 1960), the linked activity of these two systems could be explained in terms of their anatomical situation, which could admit of a common system of innervation.
Only a few preparations of Stomphia were vitally stained with methylene blue. Bipolar nerve cells, much like those described for other anemones (Hertwig & Hertwig, 1879; Pantin, 1952), are present at least in the mesenteries (Pl. 2f). In the column, however, there are histological elements resembling multipolar nerve cells (Pl. 2e). If multipolar nerve cells are present in Stomphia, it is possible that they may be concerned with initiating or maintaining excitation in the parieto-basilar system during swimming.
RECOVERY
Detailed observation of recovery after swimming did not form part of the present study. The anemone comes to rest on its side and is relatively unresponsive to Dermasterias and other stimuli. The column becomes shorter, and as the tone of its circular muscle is relaxed, that of the disk and tentacles returns to normal. Yentsch & Pierce (1955), and Sund (1958), estimate that the anemone ‘re-establishes its hold on the substratum’ after 1 or 2 min. The normal sessile position may be regained with corresponding speed, but in some cases the whole recovery phase lasts longer. Once the anemone has reattached, it will usually respond to stimulation with Dermasterias by swimming again.
At the stage when the pedal disk is about to reattach to the substratum it becomes extremely adhesive. The surface is thrown into fine corrugations and will fasten immediately to any solid object. The adhesion is such that the whole anemone may be lifted about on the tip of a mounted needle. The secretion of an adhesive substance, while probable (see layer of cuticle below the epithelium of attached specimen, Pl. 2d), would seem insufficient to account for this, and local contraction at least of the epithelium is probably involved as well (see p. 352). In an anemone lying on its side after swimming the edge of the pedal disk nearest to the substratum adheres to it, and contact is rapidly regained over the whole disk.
DISCUSSION
The ‘column system’ is activated by an initial response of the tentacles, sphincter, and possibly retractor muscles to Dermasterias. All observations so far suggest that whereas the rapid sphincter-retractor system can excite or inhibit the column, the converse does not hold. A swimming anemone, for example, will retract if given a sharp prod, and the column system is inhibited until the sphincter and retractors relax, when swimming is usually resumed. A one-way relation between the sphincterretractor and column systems may possibly occur in Metridium, since events in the rapid through-conduction system may either excite or modify the activity of the column, but not vice versa (Batham & Pantin, 1954).
Although the swimming movements of the column are the most striking part of the whole response, it is not yet clear how they are propagated. Assuming that excitation is maintained in the column, there are two possibilities. Either each bending movement in the series depends on excitation set up by the preceding one after the manner of a chain reflex, or it is fired off by one or more pace-makers. There is little evidence for the first possibility except in the particularly short interval between two contractions when they are at diametrically opposite radii (p. 353). Here it seems as though the second contraction of a pair may be a direct result of the first. The simple explanation which would be provided by a sensory reflex is not very satisfactory. It could be supposed, for example, that every parieto-basilar contraction causing the anemone to bend excited the sense cells at the opposite radius of the column mechanically, by stretching. Sense cells in the column are endodermal, and they are aggregated along each side of the mesenteries at their junctions with the column (Pantin, 1952 ; Batham et al. 1960). It has been shown in Calliactis that they respond particularly to stretch and to other mechanical deformation tangential to the surface of the column. But existing evidence does not support the idea of this kind of ‘stretch reflex’ in Stomphia. Paired parieto-basilar contractions have been observed in experimentally incised anemones which were so little expanded, owing to the loss of coelenteric pressure, that no part of the column could possibly have been stretched.
The possibility remains that every parieto-basilar contraction stimulates the sense cells along its own radius. This excitation would in turn have to produce a contraction on the other side of the anemone. If excitation were conducted in both directions round the circumference from the first site of contraction, a symmetrical position of the second parieto-basilar contraction would follow. On the other hand, the third contraction very rarely occurs at the same radius as the first, and it is preceded by a longer interval than that between the first two contractions. Even allowing for refractoriness, inhibition, or fatigue in some part of the neuromuscular system at the first radius of contraction, no simple theory of sensory reflexes can fully explain the contraction sequence of the parieto-basilar muscles.
To whatever extent sensory reflexes may be involved in the pattern of swimming movements, the nervous system plays an important part in this activity. It does not seem likely that chain reflexes of nervous, any more than of sensory origin, could alone account for a series of parieto-basilar contractions lasting for several minutes, and some kind of pacemaker system most probably exists. This might be localized in one region, or else consist of several active sites, which would not necessarily occupy fixed positions. It is possible, as suggested above (p. 357), that one or several pacemakers are situated in the column. Whether or not this can be confirmed, much more information about the neuromuscular system of Stomphia is required before a satisfactory explanation for the type of record shown in Text-fig. 10 can be put forward (see Hoyle, 1960).
There are few records of swimming in other actinians. Planktonic forms (e.g. Minyas) float passively, and Aiptaisia(Portmann, 1926) can creep slowly by performing burrowing movements when in a horizontal position. The only published description of rhythmical activity as rapid as in Stomphia seems to be Verrill’s note (1928) on Nectothela lilae.* This is a small Hawaian anemone related to Anemonia, which possesses long and extremely numerous tentacles. It can swim actively, apparently by rhythmical contractions of the disk, during which time the column is markedly shortened. Here a temporary pacemaker is almost certainly present. Carlgren referred Nectothela to Boloceroides mcmurrichi and found a similar species in the Red Sea (1899, 1927, 1949), and it is also recorded from Japan (Uchida, 1938). Gonactinia proliféra can also swim, apparently in the same way (Stephenson, 1935).
Although there is little information about the distribution of the nervous system and receptors in Stomphia, actinian nerve cells usually run between muscle fibres and the epithelium, and thus in general follow the plan of the muscles (Batham et al. 1960). Sections of Stomphia show that neurites in this situation could easily link up the responding muscle systems without passing into the mesogloea, and thus transmit excitation between them. Future histological studies may be expected to reveal nervous connexions between the tentacles and sphincter; between the sphincter and the retractors on the one hand, and the circular/parieto-basilar system of the column on the other; and between this last system and the pedal disk. In the absence of more direct information, inspection shows that anatomical junctions between the different muscle systems are sufficient to account for most of the pathways of excitation which are indicated by the animal’s swimming behaviour.
The effectiveness of actinian muscle systems depends upon the hydrostatic skeleton provided by sea water in the coelenteron (Batham & Pantin, 1950a; Chapman, 1958). In this respect, Stomphia may be regarded as an almost watertight cylinder, for the mouth is the only opening (p. 345). The mesenteries partition the coelenteron so efficiently that vertical halves of a cut anemone still maintain approximately normal tone instead of collapsing. Coelenteric pressure has not been measured in Stomphia, but in Metridium it averages 2− 3 mm. H2O (Batham & Pantin, 1950a). Bending movements of the column during swimming depend on the turgor of the crown region, which is maintained by circular tone of the column. The parieto-basilar muscles, bending the crown towards corresponding radii, act through mesogloeal ‘tendons’ in the upper regions of the mesenteries (p. 346). In anemones incised for experimental purposes, the crown may fail to expand and parieto-basilar contractions then simply produce grooves in the disk.
The initial closure of the anemone forces out coelenteric fluid through the mouth. In the rapid expansion which follows, water is taken in again so quickly that ciliary action of the siphonoglyphs would hardly seem adequate. In Metridium, Batham & Pantin (1950 a) found that anemones recovering from extreme contraction could show negative coelenteric pressures. Their suggestion that suction might be exerted by peristalsis would also apply to Stomphia. In one specimen, the over-all specific gravity could decrease from 1· 08 to 1· 06, implying a significant increase in buoyancy during swimming. In natural conditions, a swimming anemone would be carried much further by currents than its own random movements might suggest. Sund’s underwater observations bear this out, as one specimen travelled m. while he watched it.
The response of Stomphia to certain starfish has parallels among some of the associations between other marine animals. Several commensals, for example, show behaviour preferences for host-specific chemical substances (Davenport, 1955). Perhaps the best-known active response to starfish is the clapping reaction of the scallop Pecten, which can hardly be interpreted as other than an escape mechanism. Lecomte (1953) found that both mechanical and chemical stimuli were needed, and that the latent period of the response varied according to the species of starfish used. Bullock (1953) has reported the flight reaction of the gastropod Nassarius induced by starfish, and Clark (1958) has shown that herbivorous littoral gastropods will move rapidly away in the presence of certain carnivorous ones.
The action of reduced glutathione on Hydra(Loomis, 1955) provides an example of prolonged activity induced by a known chemical substance. In suitable concentration it evokes a complete feeding reaction. The gastrozooids of the siphonophore Physalia react in the same way (Lenhoff & Schneiderman, 1959), but the substance is without effect on several species of sea-anemones, including Stomphia(Passano, 1957). The responses of anemones to chemical stimuli may be seen, however, in feeding (Pantin & Pantin, 1943) and spawning behaviour (Nyholm, 1949). In general, dissolved chemical substances affect at least the slow, phasic activity of the column (Batham & Pantin, 1950c), and often that of the crown and foot as well.
The swimming movements of Stomphia differ in their rapidity, however, from the much slower activities of anatomically comparable regions in other sea-anemones. In Metridium, walking, for example, is a complex activity involving the pedal disk and lower column (Parker, 1919; Pantin, 1952). It is extremely slow compared to the rapid reactions of the pedal disk in Stomphia, but both can take place in transected lower halves of anemones, and the neuromuscular co-ordinating mechanisms may be similar. Some approach to the localized parieto-basilar response of Stomphia is found in the Brazilian anemone (?Bunodactis) studied by Pantin & Vianna Dias (1952). It responds to gentle prodding by rapid flexion of the column. This is much faster than the slow parietal shortening which is produced in most anemones by local illumination and other stimuli. Time-lapse films show that when much expanded in the presence of food., Metridium may also perform slow bending movements, in which the animal sways, of flexes at the base of the column. Batham & Pantin (1950a) interpreted the latter as a tendency for the anemone to overbalance when its coelenteric pressure is very low, and these sweeping movements do not seem to have much in common with parieto-basilar contractions.
It seems possible to derive various components of Stomphia’s swimming response from ordinary patterns of behaviour in other actinians, but this throws no light on how such a response could have evolved, especially as its biological significance is still obscure. Specialized anatomical features in Stomphia reflect its ability to swim, and it is possible from Carlgren’s record (1921) of the allied deep-water Actinostola spets-bergensis that other anemones in this family may also swim.
Of the two starfish known to cause swimming in Stomphia, Dermasterias imbricata is the only species of its genus, and is strictly indigenous to the Pacific coast of North America (Fisher, 1911). Hippasteria spinosa has a similar distribution, but other species occur further north, and H. phrygiana ranges from Scandinavia to North America across the Atlantic. It is thus possible to correlate the distribution of Hippasteria with that of Stomphia to a certain extent (see p. 343). 5. coccinea dredged off the north coast of Denmark gives a strong swimming reaction to specimens of H. phrygiana which occur occasionally in that area (Dr Gunnar Thorson, personal communication). According to Mortensen (1927) H. phrygiana occurs at depths between 20 and 800 m., and feeds on echinoderms, mussels and worms. If Hippasteria fed on Modiolus, which seems to be a fairly common substrate for Stomphia, it is just possible that the anemone might swim as an escape reaction. On the other hand, if live specimens of Stomphia can be obtained outside the range of Dermasterias and Hippasterias, and different animals are found able to excite a swimming reaction, the problem will present yet another facet of peculiar interest
ACKNOWLEDGMENT
It is a pleasure to thank the U.S. Office of Naval Research (Biology Branch) for financial support which enabled this work to be carried out; Dr D. L. Ray for the privilege of being much indebted to her; the Director of the University of Washington Marine Laboratories at Friday Harbor and his staff and colleagues for their assistance; Dr H. Fraser-Rowell for taking photographs for Pl. IA and B; and Prof. C. F. A. Pantin, F.R.S., and Dr E. W. McConnachie for helpful criticism of the manuscript.
REFERENCES
EXPLANATION OF PLATES
Plate 1
Anaesthetized specimens fixed in picroformol.
Vertical half of specimen with crown only partly relaxed.
A perfect mesentery, with its attachment to pharynx, disk, body wall and foot. Note especially the vertical folds of the retractor muscle (endocoehc), and the triangular parietobasilar muscle (exocoehc) whose border is shown by the arrow. Part of the opaque region of the mesentery which continues upwards from this muscle (top left) is a thickened strip of mesogloea (see Text-fig. 2).
(c), (d), (e) Surface views of thick slices taken at approximately the levels shown by the arrows in (a).
Plate 2
From a transverse section of sphincter (Susa-Mallory), showing muscle fibres lining cylindrical spaces in the mesogloea. The material is poorly preserved.
Transverse section showing the junction of two perfect mesenteries with the column (to the left). Note endocoelic retractor muscles, and the parieto-basilars, developed as a fold on the outer surface of each mesentery. The apparent split in the mesogloea of each septum indicates the position of exocoehc radial muscle fibres running in a transverse direction (Susa-Masson).
Vertical section of the foot along a radius. The ectodermal epithelium is just seen at the bottom of the picture. Between the oval cross-section of bundles of circular muscle fibres embedded in the mesogloea obhque spaces run towards the epithelium. The fibres lining these spaces are those of the parieto-basilar muscle, which thus make contact with the surface of the foot (Susa-Mallory).
Highly magnified portion of epithelium from the foot (mesogloea at top of photograph). Note numerous vertical strands which have stained with acid fuchsin, their blunt terminations in the mesogloea, and the cuticle covering the surface (Susa-Mallory).
(e), (J), (g) From preparations fixed after vital staining with reduced methylene blue, (e) A multipolar nerve cell from the mid-column region of the body wall. (f) Central part of a bipolar nerve cell from the tractor face of a perfect mesentery, (g) Sense cell with two neurites, and the flagellum just visible the arrow, from the pedal disk.
Mr R. Josephson kindly drew my attention to Nectothela