Experiments On Choice of Substrata by Spirorbis Larvae (Serpulidae)

At Swansea there are three closely related species of Spirorbis occurring each on a characteristic substratum: S. borealis Daudin on Fucus serratus, S. corallinaede Silva and Knight-Jones (1962) on Corallina officinalis and S. tridentatus Levinsen on rocks and stones. This pattern in their natural distribution suggests a high degree of substratum selectivity by the larvae at the end of their pelagic stage. It is known from previous studies that larvae of S. borealis tend to select various substrata, settling readily and abundantly in the laboratory on F. serratus, F. vesiculosas, F. platy carpus and Laminaria saccharina but sparsely on Ulva lactuca, Pelvetia canaliculata, Ascophyllum nodosum, Himanthalia lorea, RJtodymenia palmata, ascidians, shells and stones (Gar-barini, 1936; Gross & Knight-Jones, 1957).

To investigate the settling behaviour of S. corallinae and S. tridentatus, adults of the various species were collected from the shore at Mumbles Head and Bracelet Bay, Swansea, during the summer and generally during periods of neap tides, when larvae are most common (de Silva, 1962). Many of the adults were incubating embryos, some of which would hatch readily if the adult tubes were broken in the laboratory. Larvae obtained in this way seem to be just as discriminating during their search for a substratum as are those which hatch without artificial aid (de Silva, 1958).

The method followed was similar to that used by Knight-Jones (1951). Crystallizing dishes, wiped clean and containing about 100 ml. of sea water were placed in narrow boxes painted white inside. These boxes were kept near a window with their long axes directed towards it so that the two sides of each dish were equally illuminated. The experimental substrata were placed on opposite sides of each dish, the positions being reversed in alternate dishes, to compensate for any small undetected differences in illumination. Varying numbers of freshly liberated larvae were added to these dishes.

The first two series of experiments involved larvae of S. borealis. In the first series these were offered a choice between F. serratus and Corallina officinalis (Table 1). In the second (Table 2) the choice was between F. serratus and a small stone which had been kept in sea water for a few days to allow it to develop the film of micro-organisms without which stones are very unfavourable for the settling of this species. Of the three substrata offered to S. borealis larvae in these experiments, F. serratus was the most favoured, filmed stones being somewhat less favourable and C. officinalis very unfavourable. The total number of larvae settling on C. officinalis was less than 2 % of the total number settled. However, this alga was chosen most loyally by larvae of S. coraltinae (Table 3), whereas F. serratus was extremely unfavourable to this species, the number settling on this alga being less than 5 % of the total number settled. The comparatively poor settlement recorded in Table 3 was partly due to these experiments having been carried out under bright light. It was found subsequently that better results could be obtained by using dim light. The larvae continued to select Corallina, however, no matter how the illumination was varied.

In similar experiments with S. tridentatus larvae (Table 4) all the larvae settled on filmed stones, showing that both F. serratus and C. officinalis are very unfavourable for the settling of this species. They will, however, settle to some extent upon these algae if no other substratum is available. Indeed, when offered a choice between F. serratus and an unfilmed stone, which had been freshly immersed in sea water, more settled on the Fucus than on the unfilmed stone. This avoidance of unfilmed surfaces may help to prevent them from settling on scoured or abraded surfaces where they would probably soon be damaged by further scouring.

In their natural habitats a variety of other substrata are available to the larvae. In order to simulate more closely these conditions further experiments were carried out with the help of a turn-table apparatus of the sort used by Crisp & Ryland (1960). In this all the experimental substrata are placed round the periphery of a single dish, which is rotated slowly. The substrata offered in each dish were of similar size and the rotation ensured that they were all similarly illuminated. The results (Table 5) indicated that F. serratus, Ulva lactuca and filmed stones are most favourable for the settling of 5. borealis whereas other algae, including F. vesiculosas and C. officinalis, are unfavourable.

In natural conditions S. borealis is much more common on F. serratus than on U. lactuca. It therefore seemed desirable to carry out further experiments involving choice between these two algae. The results of these showed clearly that S. borealis larvae choose F. serratus in preference to Ulva lactuca (Table 6).

Further experiments with S. corallinae larvae (Table 7) showed that C. officinalis on which this species typically occurs is the most favoured substratum, Chondrus crispus and filmed stones being much less favourable. On certain shores S. corallinae occurs on both C. crispus and Gigartina stellata but it has not been observed so far on rocks and stones.

A final turn-table experiment (Table 8) with larvae of S. tridentatus showed that filmed stones, on which this species normally occurs, are the substrata most favoured by its larvae, algae being totally unfavourable.

It is therefore probable that these three species of Spirorbis occur on their typical substrata largely because of the high level of discrimination shown by the larvae during settlement.

On the shore S. tridentatus is restricted to dimly lit places such as gullies, crevices, deep tide pools and the sides and roofs of caves. Since S. borealis and S. corallinae occur mostly on algae they are necessarily found where the light is stronger. It seemed desirable to investigate whether there were different responses to light by larvae of the different species, which might bring them to the zones occupied by the algae on which they are characteristically found or to the deep crevices, etc., where algae are absent.

In experiments to study this possibility a square cardboard box was used, partitioned into sixteen chambers of similar size and blackened inside. Its lid had an equal number of circular apertures, each corresponding to a chamber in the box and somewhat narrower than the chamber. Films of micro-organisms were allowed to accumulate for a few days in beakers containing sea water, both on the sides of the beakers and on the interfaces between water and air. These films provided favourable surfaces for the settling of the larvae. Each of the beakers was kept in a separate compartment and the entire box was illuminated from above, but the illumination in each compartment was varied by covering the apertures with neutral filters of different opacities (Fig. 1). Various numbers of larvae were added to each beaker and after a few hours the numbers settled above and below half-depth were counted (Table 9).

These experiments showed that the brightness of the light, within the rather narrow range used, is unimportant and that S. tridentatus is different from S. borealis in that a higher proportion seeks deeper levels for settling when illuminated from above. If there is no light both species, particularly S. tridentatus, are positive to gravity. Nevertheless, S. tridentatus is very common on the roofs of shore caves, as though the larvae of this species swim upwards if the light is from below. To investigate this possibility further experiments were carried out with the partitioned box upside down, and the beakers illuminated from below. The larvae of S. tridentatus then tended to swim upwards, those of S. borealis downwards (Table 10).

Evidently larvae of S. borealis and S. tridentatus show different responses to light during settling. Those of S. borealis tend to be positive to both light and gravity. Those of S. tridentatus are still more positive to gravity in the absence of light and are overridingly negative to light. Under natural conditions the larvae of S. tridentatus must be prevented from settling on algae not only by a direct choice of a stony sub stratum but also by a preference for dimly lit places where algae cannot grow.

S. borealis, S. Corallinae and S. tridentatus are so closely related as to have been confused by previous investigators. Although they can be distinguished by small morphological differences, the most striking differences between them are the substrata which they favour and the chief evidence justifying their separation as three good species is provided by these experiments on their settling behaviour.

It seems clear that the ability of these larvae to select the substratum characteristic of each species is one that is genetically determined and not the result of conditioning during the embryonic or larval development, as may occur in insects (Thorpe, 1939). Knight-Jones (personal communication) has observed that the larvae of S. borealis on Ascophyllum nodosum (for this species occurs very rarely on this alga) choose in laboratory experiments F. serratus in preference to the substratum on which their parents occurred. Similarly, larvae of this species from parents occurring on Laminaria hyperborea in Milford Haven were found to select F. serratus in preference to L. hyperborea.

The chief criterion for separating species is that continued interbreeding should not take place between them in the wild state. It is probable that Spirorbis, like most sedentary marine polychaetes, liberates spermatozoa into the surrounding sea water. As the three species studied here commonly occur together in the same locality, often in different parts of the same small tide pool, it would appear that the spermatozoa of one species may be carried by the water to the vicinity of another. Unfortunately, it proved impossible to obtain laboratory evidence on whether or not these species can hybridize. Extensive and time-consuming experiments, using Phaeodactylum as food, failed to get Spirorbis to breed, when isolated in pairs in small (50 ml.) beakers. Field observations, however, suggest that there is no interbreeding between these forms in the Swansea area. Many thousands of worms belonging to the three species were examined during this investigation, without discovering any that could be considered hybrids.

In Milford Haven there seems to be a more problematical situation. A form is common there which is in some characters intermediate morphologically between S. borealis and S. corallinae, but which chooses to settle on F. serratus rather than on Corallina. It is possible that hybridization between these two species has taken place in the extraordinarily sheltered Haven. Perhaps some substratum favourable to both, and thus allowing free mixing of populations, occurs commonly in that area. Further investigations are needed in Milford Haven and other unusual localities. But in collections from more typical rocky shores, ranging from Plymouth to Anglesey, all three species were abundantly represented, with no evidence of hybridization. They seem more distinct than the genetically determined ecotypes of Nucella (Staiger, 1957).

It has been known for some time that pelagic larvae of various marine animals are careful in choosing a substratum and can postpone metamorphosis if a suitable substratum is not found (Thorson, 1950; Wilson, 1952). Cyprids of the barnacle, Balanus balanoides, are stimulated to settle by contact with the epicuticle of barnacles of their own species and not by those of other species (Knight-Jones, 1955). Such specific reactions are probably found in the settling behaviour of many marine larvae. Gregariousness during settlement has been demonstrated in several forms, including S. borealis and S. pagenstecheri (Knight-Jones, 1951). The two species S. corallinae and S. tridentatus show a similar tendency. Gregariousness may contribute towards reproductive isolation of populations occupying restricted habitats, by reinforcing the tendency to choose such habitats and by providing greater opportunities for inbreeding within the populations thus established.

However this may be, marked differences between strains in choosing substrata must tend to bring about ecological isolation, reinforcing other barriers to interbreeding and helping to perpetuate differences that may have arisen during periods of geographical isolation. It is possible, indeed, that the species dealt with here, like those studied by Diver (1940), may never have been separated by an absolute barrier. Millicent & Thoday (1960) have shown that evolutionary divergence can occur between populations which are not completely separated. It may have been diverging trends in settling behaviour, leading to gradually increasing skill in the choice as substrata of certain dominant algae, that in due course virtually restricted certain strains to those algae and thus contributed to speciation in Spirorbis.

Another closely related form which may have evolved in this way is S. rupestrisGee & Knight-Jones (1962), but though this is abundant at Plymouth and elsewhere it does not occur at Swansea and thus escaped my notice.

I am very grateful to Prof. E. W. Knight-Jones and Prof. G. E. Newell for advice and helpful comments.