Laboratory Experiments on Gregariousness During Setting in Balanus Balanoides and Other Barnacles

Cirripede larvae must become widely dispersed during their long planktonic life, yet the adults of many species are found exclusively in restricted habitats. Some are epizoic and others are parasitic on specific hosts, which the cypris larvae must therefore seek out land recognize, whilst cyprids of many Lepadidae and Rhizo-cephala are known to seek out adults of their own species, to which they attach themselves as dwarf or complemental males. Burton (1949) pointed out that cyprids of Balanus balanoides (L.) seem to be gregarious during setting, and field experiments indicated similar behaviour in Elmimus modestus Darwin (Knight-Jones & Stevenson, 1950). The following experiments have established that these common littoral forms are able to recognize their own species, and have thrown some light on the mechanism of recognition.

Cyprids of Balanus balanoides were collected near Bangor during April and May, by picking up Fucus from shallow water and rinsing the fronds in wide-mouthed jars (breffits), which contained about 2 1. of sea water. Large numbers of cyprids were dislodged from temporary attachment to the fronds, and were later pipetted away to clean vessels, taking advantage of their positive phototaxis.

Cyprids of Balanus crenatus Bruguière and Elmimus modestus were obtained from plankton hauls. They could be readily separated from the rest of the catch, as they became trapped in the surface film when left undisturbed in shallow dishes. Apparently the cyprid cuticle is more strongly hydrophobic in these species than in Balanus balanoides.

Experiments were carried out in uniformly illuminated glass dishes, containing 250 ml. of sea water, advantage being taken of the fact that glass is unfavourably smooth for the setting of barnacles, especially of B. balanoides (Runnström, 1925; Barnes, Crisp & Powell, 1951). Records were made of the numbers setting on surfaces which had been treated in different ways. Cyprids were usually active, alternately swimming and crawling. Their exploratory habits enabled them to choose between two surfaces presented to them in the same dish. A cyprid which was about to settle could be recognized because it crawled slowly and turned frequently (Doochin, 1951), as do the larvae of Ostrea (Cole & Knight-Jones, 1939) and Spirorbis (Knight-Jones, 1951), when they are about to settle. After setting, no further locomotion was possible, and a settled cyprid could easily be recognized by its firm attachment and obtuse anterior end. The interval between setting and the moult which accompanied metamorphosis was about 24 hr. in Balanus balanoides, somewhat less in B. crenatus, and less than 4 hr. in Elminius modestus.

It was repeatedly shown that cyprids did not settle readily except on surfaces which already bore either barnacles of their own species or the persistent bases left after these barnacles had been removed. The following experiment proved that failure to settle was not due to development being insufficiently advanced, and that setting could be delayed for long periods without ill effects. A hundred and forty freshly collected cyprids of Balanus balanoides were divided equally between seven dishes, and a Mytilus valve bearing the bases of freshly detached barnacles was placed in one of the dishes on the following day. Similar valves were added, at intervals of 4 days thereafter, to each of the other dishes in turn, and notes were made every day of the condition of the cyprids and of the numbers settled on the valves (Table 1). None settled on the glass of the dishes.

In the first dish to receive a Mytilus valve the majority of the larvae settled within 24 hr., and it is therefore probable that the majority of those in other dishes would have settled within this period if a suitable substratum had been available. Where none was available the planktonic life was prolonged by nearly 2 weeks without serious ill effects, for most cyprids presented with a valve after this period were still able to settle within 24 hr. and to metamorphose. These metamorphosed individuals, though somewhat below the normal size, seemed healthy and started to grow, but after further prolongation of the planktonic life the majority of cyprids became sluggish, encumbered by bacterial growths, and apparently unable to settle.

Seven parallel series of ten experiments were set up, each experiment with a similar batch of twelve B. balanoides cyprids and one Mytilus valve, from which the flesh had been removed. In three series of experiments the Mytilus valves bore live individuals of B. balanoides, B. crenatus or Elminius modestus respectively. In another three series they bore bases only of these species, and in the last series they bore no trace of barnacles (Table 2A). The valves were all similar and had been collected in such a way as to ensure that the absence of Balanus balanoides, from those which were bare or which bore other species, was due to geographical distribution, rather than to the valves being intrinsically less suitable for setting. Thus some had been collected very early in the season from Cardigan Bay, and these bore numerous newly settled B. balanoides ; whilst the others, which had been collected at about the same time from Bangor, Conway and Rhyl, bore no B. balanoides because it was too early in the year for setting of this species to the east of Anglesey, where liberation of larvae regularly occurs later in the season (Crisp, 1953). The experiments showed that the other species of barnacles were comparatively ineffective in promoting setting, though they appeared to be better than nothing. The fact that more cyprids settled on valves bearing B. crenatus, than on those bearing Elminius modestus, may possibly have been associated in some way with the closer relationship of the former to Balanus balanoides. Alternatively, it may have been due to the cyprids reacting to surface irregularities, since the individuals of B. crenatus were larger and more numerous than those of Elminius modestus, whilst their calcareous bases remained on the valves after the barnacles had been removed, making the surfaces rougher.

Similar experiments, with similar results, were carried out with pieces of sandstone from Brixham, Devon, which bore either Balanus balanoides, B. perforatus Bruguière, Chthamalus stellatus (Poli), or Verruca stroemia(O. F. Muller) (Table 2 B). Fewer cyprids settled in these experiments because the adult Balanus balanoides were few, whilst the stones which bore them were so crowded with Chthamalus as to leave little space for setting.

Experiments in which cyprids of various species were offered a choice between two Mytilus valves placed in the same dish, one bearing bases of their own species and the other bases of another species, showed that the larvae can select surfaces bearing their own species, in preference to those bearing another (Table 3). The experiments with Elminius cyprids, which were carried out at Bumham-on-Crouch, involved choice between Mytilus valves from Rhyl, which bore Elminius and a few Balanus, and similar valves from Bangor, which bore Balanus only, because Elminius had not spread as far as Bangor. Some of the valves had been stored in 5 % formaldehyde for 4 days, between the time of collection and the time of the experiments, and it is interesting to note that the specific effect was unaffected by this treatment (cf. p. 594).

The positive results of these experiments on choice indicated that reception of the stimulus to settle requires close proximity to, or contact with, the previously settled individuals or their bases. They further suggested that cyprids which have received this stimulus rarely swim away to another surface before setting. Subsequent experiments (p. 591) confirmed that contact was necessary for the reception of the stimulus, and repeated observations showed that cyprids which came into contact with barnacles or bases embarked immediately upon the turning movements which characterize the setting action pattern. Their subsequent migrations were usually limited, as is illustrated by the experiments described in the next section.

A slate which bore a small patch of adult Balanus balanoides (Pl. 16, fig. 1) was taken from a shore before heavy settlement of barnacles had begun there. The position of the patch was outlined by a scratched groove, and the barnacles were removed, leaving only the almost invisible bases. The slate was left overnight, exposed to cyprids in a large vessel, lying horizontally and illuminated from above. The cyprids, which started crawling over the whole surface of the slate, settled almost exclusively within the area outlined (Pl. 16, fig. 2) which suggests that little or no migration occurred after contact with the bases. A similar experiment was carried out with a slate which bore B. crenatus as well as B. balanoides to illustrate the ability of the cyprids to distinguish their own species (Pl. 16, figs. 3, 4). In this experiment the positions of the barnacles were not outlined before removal, because they were rather scattered, and because the calcareous bases of B. crenatus were fairly conspicuous, whilst the relative positions of the other bases were known from the photograph of the intact barnacles. Comparison of figs. 3 and 4 (Pl. 16) shows that few cyprids settled in the areas which had been occupied by B. crenatus, many in those which had been occupied by B. balanoides, and many in grooves and depressions nearby, to which they had presumably crawled.

It might be expected that cyprids encountering barnacles on surfaces which are otherwise unfavourable for setting would receive the stimulus to settle, but would swim off during their final reconnaissances, and become pioneer colonizers of neighbouring bare surfaces. An opportunity to demonstrate this occurred when two dishes, in which experiments had been carried out, were each found to contain a recently metamorphosed B. balanoides attached to the glass. Very few cyprids of this species (much less than o-i %) settled on bare glass in the whole course of this work, but once these abnormally undiscriminating individuals had begun colonization of the glass surfaces others settled beside them when successive batches of numerous cyprids were introduced into the two dishes. After about twenty had settled in each dish, the majority of the cyprids which were introduced attached themselves to the adjacent glass within 2 or 3 days, there being no other surface available. When each dish bore about fifty settled barnacles, the two dishes were matched with two other dishes which bore none. Similar batches of cyprids were added to all four dishes, and similar stones of rough granite were suspended freely in the water by cotton threads. Cyprids which had encountered the barnacles in the experimental dishes could reach the stones only by swimming, yet setting occurred much more readily on these stones than on those in the other dishes (Table 4). Contact with the settled barnacles appears to have provided a sensitizing stimulus, so that subsequent contact with the rough stones induced the final response. Setting of barnacles normally involves responses to multiple stimuli, being delayed by crowding and orientated by light and currents (Crisp, 1953), and the extent to which setting is localized, after contact with adults, evidently depends upon the presence of a favourable surrounding surface.

To see if any part of the body other than the basis would induce setting, pieces of adult barnacles were placed on top of similar freshly quarried granite stones, which were offered to cyprids of B. balanoides in four parallel series, each of twenty experiments. In a parallel control series (a) there were no fragments on the stones. In the four experimental series respectively, each stone bore (b) two detached adults of B. crenatus’, (c) two detached adults of B. balanoides, complete except for their bases; (d) the bases of about twenty detached adults, scraped with a scalpel off smooth pebbles or shells, and scattered on the granite in small, flocculent heaps; (e) the bodies of three adults extracted from their shells, including the cirri and viscera (Table 5 A). No cyprids settled on the untreated stones, but significant numbers settled on those which bore fragments of their own species. A few cyprids settled beside the adults of B. crenatus, giving further slight evidence that setting occurs more readily in the presence of barnacles of other species, than in the complete absence of previously settled barnacles (cf. p. 586).

The cyprids setting alongside the adult cirri and viscera, with which they do not normally come into contact, indicated that setting occurs in response to some substance which is generally distributed throughout the body. Since bases alone promote setting, the active substance is probably cuticular, for the attachment cement which largely constitutes the basis is similar in properties to the cuticle, and the cement glands are regarded as modified tegumental glands (Thomas, 1944). Cyprids have relatively large cement glands communicating with the antennules, and when dissected with needles they yield considerable quantities of a sticky substance. In a series of experiments slates which had never borne barnacles were broken into paired pieces, and one piece of each pair was presented without treatment to a batch of cyprids, whilst the other was presented to a similar batch after having been smeared thoroughly with from ten to twenty dissected cyprids, left for a few hours in sea water and then rinsed. Twenty-four hours later none of the cyprids had settled on the untreated slates, but in most batches several had settled on slates smeared with squashed cyprids (Table 5B). It is not certain that they were responding solely to the extruded cement, since fragments of the cyprids’ bodies were inevitably mixed with this.

The numbers of cyprids setting in these experiments (Table 5 A and B) were comparatively small, and this reluctance to settle could hardly have resulted from a phobic response to damaged tissues, because placing damaged barnacles amongst attached adults did not in the least discourage setting. Cyprids reacted readily only to cuticular material which was firmly attached to a surface, presumably because it was more readily encountered by their groping antennules, or perhaps because looseness of a surface discourages setting. When kept crowded for a long time they often clung to one another in temporary aggregations, but such contact did not make them more ready to settle, which suggests that they respond only when crawling.

The following experiments were carried out to test whether the gregarious response depends on chemosensory perception of a water soluble substance emanating from settled barnacles, or whether contact with the barnacles is necessary. Small shells or stones bearing many settled barnacles were enclosed in bags of bolting silk, to prevent cyprids from touching the barnacles, and each bag was placed in a dish containing a batch of cyprids and a Mytilus valve. These dishes were grouped in two series of experiments, one series with Mytilus valves bearing bases of freshly detached barnacles, and the other with valves bearing no bases. Setting readily occurred upon the valves which bore bases, but very few cyprids settled upon the others; indeed, not significantly more than in a third series of experiments in which valves without bases were offered to cyprids in dishes without bags of settled barnacles (Table 6 A). It appeared that there was nothing to encourage setting in the water which surrounded the settled barnacles.

In further experiments, setting was prevented by covering barnacle bases with a thin but continuous film of nitrocellulose, by painting with 2 % celloidin in etheralcohol. After drying under reduced pressure setting occasionally followed, but subsequent examination showed that the film was cracked in many places, and that the cyprids had settled in the cracks and on the celloidin adjoining them. When the film was hardened by immersion in 90% alcohol, and then washed in hot water to remove the ether-alcohol, it was entirely effective in preventing setting, though such a film would be permeable to most water-soluble substances. It may be inferred that cyprids do not settle readily, unless they have come into physical contact with settled barnacles or their bases.

The reaction of cyprids to bases enabled the following experiments to be carried out, to test the stability and chemical nature of the substance to which they respond. Pieces of slate, covered with the bases of recently detached B. balanoides, were treated in various ways and then each piece was kept for 24 hr. in a dish of sea water, containing twelve cyprids. Control slates without bases were treated similarly and exposed to cyprids under similar conditions, whilst parallel experiments were carried out with slates bearing untreated bases. Cyprids settled readily within 24 hr. on all the untreated slates, but on none of the bare control slates (Table 7G). The presence or absence of settled cyprids on the treated slates, after a similar period of exposure, therefore showed whether or not the substance which induces setting had resisted the treatment.

Batches of four slates were exposed for 30 min. to various temperatures in an electric oven, allowed to cool, washed, wiped and presented to cyprids. Setting occurred on all the slates which had been exposed to temperatures up to and including 200° C., on two out of four slates exposed to 250° C., on one out of four slates exposed to 275° C. (when charring of the bases occurred) and on none exposed to 300° C. or to higher temperatures (Table 7 A). The active substance seems to have been unaffected by heating to 200° C., but gradually destroyed at about 250° C.

The experiments recorded in Table 7B showed that the bases remained active after treatment with various solvents, which were afterwards removed by drying under reduced pressure, or by washing overnight in running tap water. Chloroform and toluene were used as fat solvents, whilst an alcoholic solution of phenol and aqueous solutions of urea and sodium sulphide were used because of their strong solvent action on some proteins and substances containing the peptide link.

Boiling in dilute acids, or treatment with cold concentrated acids, had little or no effect, but no setting occurred after treatment with boiling strong acids (Table 7 C). Setting was unaffected by cold caustic alkalis, but reduced after brief boiling in these and prevented by prolonged boiling (Table 7D).

Cold concentrated nitric acid turned the bases yellow (xanthoproteic reaction) but did not altogether prevent subsequent setting (Table 7C). The experiments recorded in Table 7E showed that setting readily followed treatment with the following reagents: pepsin, in warm acid solution (which would have attacked accessible peptide links); formaldehyde (which would have reacted with amino groups) ; benzoquinone (tanning agent) ; and freshly prepared, ice-cold diazonium solution (which would have reacted with phenolic side chains). After immersion in a solution of mercuric chloride (used on account of the strong adsorption of mercuric ions by carboxyl groups) slates were unfavourable for setting, in spite of very prolonged washing, but the cyprids became sluggish and may well have been affected by de-adsorbed mercury.

Setting occurred readily after treatment with hydrogen peroxide, but not after treatment with sodium hypochlorite, the corrosive agent which dissolves quinone-tanned proteins (Table 7F). This was observed under the binocular microscope to dissolve the bases.

A few similar experiments with bases and cyprids of Elminius modestus indicated that these cyprids, like those of Balanus balanoides, respond to a substance which is attacked by boiling alkali or cold hypochlorite, and is more resistant to acid than to alkali. The results given in the last column of Table 3 show that the activity of the substance is not affected by prolonged treatment with formaldehyde.

These chemical tests throw considerable light upon the nature of the substance to which the cyprids respond. Clearly it is not readily soluble in aqueous or alcoholic media, or in fat solvents, while the heating experiments show that it cannot be in the least volatile. These properties rule out any simple organic substance, and suggest that either a large molecule or a natural polymer is involved. This is in keeping with the specific character of the gregarious effect, which implies small variations in structure between species, such as might be expected in polymeric substances. The resistance of the substance to all but the most corrosive agents further reinforces this view. Its resistance to sodium sulphide rules out a keratin type of protein, while its greater sensitivity to dilute alkalis than to dilute acids rules out chitin. Quinone-tanned proteins, on the other hand, are known to be attacked by hypochlorites, as is found here. Microchemical tests (Knight-Jones & Crisp, 1953) have indicated that the epicuticle of barnacles, like that of other arthropods (Pryor, 19390,6; Dennell, 1947), contains quinone-tanned proteins, whilst the cement in the bases is probably of similar material (Thomas, 1944). Darwin (1851-54) remarked that this cement resembles chitin, but is more resistant to acids and less resistant to strong alkalis, and Pyefinch & Downing (1949) suggested that it is a quinone-tanned protein. It seems highly probable that the gregarious response is due to contact with quinone-tanned proteins forming the epicuticle of settled individuals. Such materials occur in many phyla (Brown, 1950), so their exact composition is likely to vary widely and may well be subject to specific differences.

It is difficult, however, to suggest a chemical basis for recognition of such substances, or to analyse the response further. It seems safe to assume that the response is not to phenolic groups, as these would be coupled and rendered insoluble by diazotization, nor to amino groups, since formaldehyde has no effect. The ineffectiveness of strong proteolytic enzymes, of nitration and of extremes of heat which would normally affect proteins profoundly may well be due to a protein framework having been stabilized and protected by a natural tanning process, but its resulting inertness would be an obstacle to recognition by any chemo-?ensory mechanism. Perhaps the antennular suckers secrete an enzyme capable of attacking it, for such an enzyme would help cyprids of parasitic cirripedes to penetrate the host’s cuticle. It is possible that barnacle cyprids respond to some system of greater than molecular dimensions, such as a pattern on the cuticular surface, but microscopical examination revealed no pattern which was common to both the bases and the parietal plates of adult barnacles.

Field experiments with Elminius modestus showed that similar bare surfaces were colonized much more rapidly on shores where adults were abundant, than on shores where there were no barnacles (Knight-Jones & Stevenson, 1950). The behaviour revealed by the experiments of Table 4 explains this effect (p. 589). Many of the cyprids, which settled on the experimental surfaces in the field, would have been stimulated by previous encounters with barnacles on other surfaces nearby, which were too crowded or too muddy to be favourable for setting.

A field experiment with Balanus balanoides was carried out at Port Dinorwic, by the Menai Straits, where a beach of slates had been formed by dumping unwanted material from a quarry wharf. Early in 1952, and for some months previously, dumping had been concentrated at one place, and the slates constituting this new area of beach bore no adult barnacles, for they had not been exposed during the setting season of the previous year. Elsewhere barnacles were common on slates, most of which had lain on the shore for many years. In May 1952 setting of B. balanoides occurred quite heavily on slates everywhere along this shore except on the new beach where there were no adults; this strongly recalled remarks by Moore (1939) on the slowness with which a new rocky shore at Plymouth became colonized by barnacles.

Eight slates covered with the bases of recently detached barnacles were prepared for the experiment, and four were exposed on the new beach, where no settlement was taking place, whilst the other four were exposed under similar conditions about 30 m. away, where barnacles were setting abundantly. Eight slates which had never borne barnacles were exposed under similar conditions alongside them. After 5 days the slates were examined and the barnacles which had settled on similar areas of each were counted (Table 8 A). Tardy colonization of the new beach had begun, for setting was becoming heavier everywhere. An analysis of variance (Table 8B) was applied to the results, which were first subjected to a square root transformation as this seemed the most appropriate in fulfilling the requirements discussed by Bartlett (1947) and Barnes (1952). The analysis indicated that the differences between slates which bore bases and those which bore none are highly significant (P < 0·001). The considerable numbers of cyprids setting on those slates bearing bases, which were exposed on the new beach, show that there was nothing unfavourable for setting there except the absence of settled barnacles. The differences between the eight slates exposed on the old established part of the beach, and those exposed on the new part, are barely significant (P < 0-05). Some difference would be expected, for many of the cyprids visiting surfaces on the old beach would have been previously stimulated by contact with barnacles on neighbouring surfaces. There was negligible interaction between the two main effects, and the interaction term was therefore included in the error variance.

Field observations for some years in Essex, Cornwall and North Wales have indicated that comparatively few larvae settle outside the zone occupied by the adults, except at times of unusually heavy settlement. At such times, when cyprids are very numerous, even isolated bare surfaces collect abnormally undiscriminating pioneer settlers, which are soon followed by gregarious individuals. The level of discrimination shown during setting seems to be subject to individual variation, and is probably lowered slightly after prolongation of the planktonic life, as in Spirorbis larvae (Knight-Jones, 1953). The normally gregarious reaction must aid breeding, since most species of barnacles appear to be cross-fertilizing (Crisp, 1950), and must also be of value in leading cyprids to habitats which have proved suitable for the survival of the species. This is particularly necessary in Balanus balanoides, which thrives only in the tidal zone.

I am indebted to Dr D. J. Crisp for most valuable advice and criticism throughout all stages of this work. As our Station was not ready for occupation, Prof. F. W. Rogers Brambell, F.R.S., and Prof. Stanley Peat, F.R.S., kindly provided facilities in the Departments of Zoology and Chemistry at Bangor. Further experiments were carried out at the Fisheries Laboratory, Bumham-on-Crouch, by kind permission of Mr R. E. Savage. I am also grateful to Dr P. N. J. Chipperfield and Miss Emily Clay, of the I.C.I. Marine Paints Laboratory, Brixham, for providing material.