Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 78, 1950
This text is also available in PDF
(3 MB) Opens in new window
– 385 –

Ecological Study of Benhamina obliquata (Sowerby),
A Basommatophorous Pulmonate in Otago Harbour

[Read before the Otago Branch, July 12, 1949; received by the Editor, July 13, 1949.]


Benhamina obliquata represents a monotypic endemic genus of the Siphonariidae. Though the anatomy has been described by Hutton (1882) and Cottrell (1910, 1911), no ecological study of this species or of other New Zealand Siphonarias has been published, and the literature on exotic species is extremely limited. The present work is therefore an attempt to investigate the habits and distribution of one species, B. obliquata. Work has been confined to Otago Harbour, though preliminary investigations at Bluff Harbour indicate that the conclusions drawn from this study may not be typical of all areas.


For breeding studies, field observations were supplemented by observation of captive specimens placed in a tidal aquarium. This consisted of an ordinary glass-sided aquarium eighteen inches in length, divided into two compartments by a vertical, loosely fitting glass partition. In either compartment, fragments of rock on which suitable algae were growing were balanced near the ends of the aquarium. Fitted on a light wooden tray and balanced on a central axle of metal, the aquarium was tilted by the action of the hour hand of an eight-day clock to which it was attached by loosely jointed levers. To take the weight of the aquarium when fully depressed, springs were attached from either end of the tank to the basal stand. The amount of sea water was sufficient when the aquarium was fully tilted to cover the seaweed-covered rocks and when tilted fully in the opposite direction to leave the same rocks uncovered. This provided conditions similar to tidal submergence every twenty-four hours. No aeration was necessary, as the surface area of water was large in relation to the depth and the movement of water underneath the glass partition served to mix the layers.


B. obliquata is found commonly in Otago Harbour wherever the rocks are large and stable and there is a supply of suitable algal food. Though they commonly collect in rock crevices, many solitary individuals are located in exposed situations.

Their position in relation to the tide appears of importance. If respiration is more efficient by lung or by gill, the species will be limited to habitats where the period of submergence will be proportional to the efficiency of each organ. To determine the relationship of Benhamina to tide levels, maps of three areas were prepared by

– 386 –

standard quadrat methods, one area being re-mapped after two years for comparison (Plates 41–15). Within these areas the numbers of B. obliquata and Siphonaria zealandica were recorded and the period of submergence calculated. (See Tables I and II.)

Table I—Period of Submergence. Benhamina obliquata
Locality Not uncovered 6 hours 5 hours 4 hours 3 hours 2 hours 1 hours Not covered
Portobello 1 23 25 2
Portobello 2 1 32 4
Cradle Rock 1 4 38 1 2
Station 1 15 7 9 6
Total 1 16 66 76 9 2
Table II—Period of Submergence. Siphonaria zealandica
Locality Not uncovered 6 hours 5 hours 4 hours 3 hours 2 hours 1 hours Not covered
Portobello 1 56 201 191 165 76 68
Portobello 2 223 192 68 96 6
Cradle Rock 1 147 72 41 49 58 3
Station 1 2 336 47 79 1 1
Total 59 907 502 353 222 133 3

Hence it is concluded that B. obliquata lives between high and mid-tide, S. zealandica between low and mid-tide, the range of the two species overlapping at mid-tide level. S. zealandica below low tide and those living above mid-tide, which were confined to rock pools where they were submerged continuously, must feed when submerged, and respiration must be wholly by use of the gill. B. obliquata above high tide mark were raised above the normal level by occupying niches where they were reached by splash. Such habitats which served to increase the normal feeding range were rare and were seldom occupied for more than a single night. If unwetted during the nocturnal high tide, these individuals did not move from their positions, and thus those which settled in such a niche would be in danger of being stranded. Several such individuals were observed to die of desiccation at Quarantine Island three weeks after the spring tides.

The distribution of both species also corresponded to the vertical zonation of the algal colonies of the harbour. Four principal zones were recognised in areas where Benhamina colonies were located:


Hormosira banksei—normal low tide.


Corallina officinalis—low to mid-tide. May be raised by the presence of rock pools.


Lichina pygmaea var. intermedia—often replaced by or associated with Symploca. Mid-tide to mean high water neaps.


Bostrychia arbuscula—mid-tide to high water springs.

Within these zones, S. zealandica was limited to the Hormosira-Corallina-Lichina zones, forming one of the dominant animals of each association; B. obliquata was found in the Lichina-Symploca-Bostrychia associations, being dominant only in the Bostrychia zone.

– 387 –

Feeding and Locomotion

B. obliquata is exclusively herbivorous, feeding on marine algae and lichens listed above.

A sharp distinction in the times of feeding and movement of young and mature Benhaminas was observed. Those of shell length under 2 cm., living at lower tide levels in close contact with algae. feed principally during daylight low tide, but they may also feed during the night. In adults, feeding is exclusively nocturnal when uncovered during low tide. Movement is spasmodic and feeding excursions may occur at intervals of several days or weeks. Abe (1935) recorded that the stimulus to motion for S. atra is the ebbing tide, and this has been found to act similarly for Benhamina. Under laboratory conditions movement may be induced by submerging specimens for about a quarter of an hour. When again uncovered, during darkness, the animals begin moving about.

Primary factors influencing movement are therefore:

(a)low tides;
(b)absence of daylight, artificial light used experimentally having no influence on behaviour.

These two factors may be modified by general weather conditions, the most important being wind, during which no movement occurs, and temperature, which operates only within wide limits. Rain had no inhibiting effect.

By peristaltic contraction of the muscles of the foot, passing rhythmically from the anterior backwards in the manner described as monotaxic retrograde by Vlés (1907), figured by Hayashi (1937), the animal is able to move over a rocky substratum at the rate of about a foot in ten minutes. The sensory head lobes are in continual contact with the substratum, the eyes are prominent and the body extended well beyond the shell margin.

The marked negative phototropism shown by adults may be overruled by negative geotropism under abnormal conditions. When placed in jars, the animals moved directly to the highest, available point, coming to rest with the sole of the foot uppermost, even though it was thus exposed to direct light. No reorientation occurred when the jars were moved unless the animals were submerged. Under water, captive specimens reoriented to the roof of the jar for periods up to a month.

Homing Habits

When experimenting on homing behaviour of Siphonaria alternata and Fissurella barbadensis, Willcox (1905) assumed that homing ability was the faculty of returning to the niche normally occupied by the animal if forcibly removed for some distance. The strength of the homing instinct was judged by the distance from which the animal returned. Other workers do not agree with this, stating that homing ability is connected with feeding, moulding of the shell, and complex and invariable behaviour pattern (Crozier, 1921).

The problem is therefore assumed to involve: (a) particular niche occupied by the animal. Recorded by Abe (1935) and Risbee (1936) for Siphonaria atra; Willcox (1905) for S. alternata.

– 388 –

Though a niche is usually marked by a scar similar to that of Patella, its presence is not in itself sufficient evidence for assuming the homing behaviour to be constant; as Willcox (1905) shows, it may be formed after the niche has been occupied for only three days. This must therefore be accompanied by:

(b) constant behaviour pattern concerned with locomotion and feeding. Recorded for Siphonaria atra by Abe (1935).

Most of the Portobello specimens were located in a large cleft running at right angles to the sea, exposed to direct sunlight only during late afternoon. From this crevice individuals emerged at low tide, moving rapidly to patches of weed over six feet away. After feeding, they usually returned to the crevice, though the same path was not always followed as Abe describes for S. atra (1935). In the crevice the original position was not usually reoccupied, most individuals settling within any part of the shelter. Those which did not return were either overtaken by the rising tide, when they would settle immediately, or would settle in temporary colonies in the rock hollows. During winter the large summer colonies frequently dispersed, reforming the following season with a mixed population of original and new members. Of 50 present at Portobello in November, 1946, 38 were marked. Of these, only 2 remained in the colony, which numbered 37 in March, 1948. The shell margins of these individuals, termed Type A, were smooth and fragile, and the body wall projected beyond the shell even when at rest.

In contrast to Type A at the higher beach levels were those termed Type B about half-tide mark. Living in exposed, solitary positions subject to wave action and direct sunlight, they possessed markedly irregular shell margins, closely fitting the rock. The place of attachment was marked by a “scar” similar to that described by Russell (1907) for Patella. From this spot the animals moved in search of food, reorienting on returning to fit the rock surface. These were observed to occupy the same “home” for three years, and the shell growth lines indicated that the niche had been occupied for some time. No shells showing a change from irregular to smooth growth lines were found. Homing behaviour was not observed in young specimens under 2 cm.

From this evidence it is postulated that there are two ecological races of Benhamina obliquata in Otago Harbour, similar to the races of Patina pellucida described by Graham and Fretter (1947). They are distinguished as follows:

Type A Type B
smooth margined shell body wall visible when resting does not occupy “home” lives between mid- and high-tide variable behaviour pattern irregular shell margin body wall covered by shell occupies “home” with scar usually lives mid-tide level or below invariable behaviour pattern

No difference in form can be observed until the animals have reached about 2 cm., when the general behaviour pattern changes. With change in habitat from mid-tide level the two forms become apparent—Type A with feeble homing behaviour which may be interpreted in terms of immediate directive stimulation, and Type B with complex and invariable behaviour pattern.

– 389 –

Age and Rate of Growth

Attempts to measure the amount of annual shell growth from marked individuals were unsuccessful because of the dispersal of the animals during the winter. An estimate by indirect methods, using the heavy disturbance lines on the shell exterior, has therefore been made. These lines, commonly known as “winter rings,” are accepted by most workers as representing periods when growth temporarily ceases during winter, but Orton (1926) points out that similar rings may be caused by disturbance during the growing period. Consequently, estimates made from these counts should be treated with reserve. The age at which the first ring is made is subject, to controversy. Orton (1926), on Cardium, takes the first winter ring as representing the end of the first year's growth, but Stephen (1931) shows this is the second winter ring. the first being so faint that it is commonly overlooked. The amount of each season's growth was measured from the winter rings where they crossed a line drawn from the shell apex to the posterior margin. (See Table III.)

Table III—Annual Shell Growth from Winter Rings, in Cms.
No of Specimen Winter Rings
1 2 3 4 5 6 7 8 9
1 1.0 1.5 2.2 3.0 3.9 4.2 4.6 5.1 5.8
2 1.2 1.8 2.5 3.0 3.5 4.0 4.6 5.4
3 1.0 1.8 2.5 2.9 3.9 4.6 5.3
4 1.0 1.5 2.0 2.8 3.6 4.1 4.4
5 1.1 2.1 2.9 3.4 3.8 4.4
6 0.8 1.4 2.0 2.6 3.6 4.2
7 0.9 1.2 1.8 2.5 3.1 3.8
8 1.0 1.5 2.2 3.2 3.7 3.9
9 0.6 1.2 1.4 1.8 2.8 3.5
10 0.8 1.3 1.9 2.7 3.3

Benhamina obliquata commonly forms the first ring at 1·0 cm. and the second at 1·8 cm. As the veligers do not hatch before November, and may be as late as March, growth to 1·0 cm. could hardly occur before winter. Hence the first ring is estimated as two season's growth. The age of the largest specimens would therefore be about ten years, but may be more, as in older shells erosion makes the winter rings less conspicuous. As breeding commences when the animals are from 3–4 cm. in shell length, this is probably their fourth or fifth season.

Growth occurs mainly in the summer. In late spring the shell margin is thin and brittle, becoming strengthened during autumn until by winter it is firm. The new growth is usually visible as a band of lighter colour round the shell margin. Table III shows that the annual growth is variable, possibly a measure of favourable and unfavurable seasons.


Congregation of the animals into large colonies in suitably sheltered rock clefts marks the onset of the breeding season. Egg coils laid in association with these colonies are earlier than those laid by solitary individuals. The colonies remain fairly constant during the summer and do not begin to disperse until autumn.

Mestayer (1920) states that the spawn coils are found in associa-

– 390 –

tion with the animals from about October until the end of April, but my observations over four seasons suggest that this breeding period, stated for the Cook Strait area, may be a little long for the southern colonies. In Otago Harbour the breeding season commences about the end of October, the different colonies laying within a few days of each other, and from then breeding is continuous until the first weeks of March.

The conditions necessary for breeding to begin have not been fully determined. Although most of the colonies start laying within a few days of each other, no correlation was found between this and the phases of the moon. Also, a large colony of over thirty mature Renhamina near Portobello, breeding prolifically in the summers of 1945 and 1946. failed to breed during 1947 and 1948, though numerous egg coils were produced by other colonies a few chains distant.

Though Hubendick (1945, 1946) states a flagellum amoratum is diagnostically not present in B. obliquata, this organ does develop on the wall of the penis, halfway down its length, at the beginning of the breeding season. This may comprise up to four flagella at different stages of development, cylindrical in form and ending distally in a sharp dart. In the preliminary courtship this flagellum is employed to stimulate the conjugants to copulate and to secure correct bodily orientation. The flagellum of the individual acting as a male is extruded through the genital pore briefly at intervals until contact with another animal is made.

During coition the conjugants assume positions with the genital apertures in contact, the animals facing in opposite directions with the heads side by side. The penis of the individual acting as a male is everted as far as the base of the flagellum, which hangs against the body wall. No movement occurs during coition, which may last several hours.

It is not known how long after fertilisation the egg coil is laid, nor whether fertilisation takes place again before subsequent coils are deposited. Commonly in mid-summer, three coils at different stages of development were found in association with single individuals on the fringe of breeding colonies. From this it would seem possible that each lays more than one coil. These are deposited during low tide, at night, but as the process is slow, may not be completed until early morning. The ribbon is ejected from the genital aperture behind the right head lobe and, when fresh, is sticky. As it is laid, the animal rotates anti-clockwise, so that a perfect spiral ribbon consisting of one to five rings is formed. When the first ring is completed, the animal mounts on to it, continuing to rotate so that the second ring is laid immediately within the first. The albumen on drying glues the ribbon to the rock and to the base of the neighbouring ring. The size and number of rings depends directly on the size of the animal. Usually two or three rings are laid, but many of four or five have been collected, the latter completely filling the coil. Each ring takes from 2–2 ½ hours to complete, the animal moving spasmodically in short jerks. Thus an average sized coil takes 6–8 hours to deposit and so must be commenced immediately the animal is uncovered by the

– 391 –

ebbing tide. One observed laying when reached by the rising tide continued without interruption, though normally the animals settle immediately.

The egg ribbon consists of creamy white coils of albumen enclosing large numbers of transparent capsules which lie in strings throughout the albuminous matrix. Each capsule contains numbers of eggs. The coils average about one inch in diameter, 10–20 in. in length and 1 cm. in height. They are distinguished from the sickle-shaped ribbons of S. australis and S. zealandica by the large size and number of rings, and from similar coiled ribbons of species of nudibranchs by the thickness of the ribbon and the numbers of eggs in each capsule. The capsules, each containing 10–30 eggs, are of clear, firm, membranous material, ovoid in shape and attached by constrictions in a continuous string. Those of S. australis and S. zealandica are similar in form, but smaller. The number of eggs, estimated by random counts along the length of the ribbon, under a squared slide, varied from 45,000 to 190,000 in each ribbon. If the supposition that each individual may lay two or three coils during the summer is correct (and the number of coils in each cleft commonly outnumbers the animals present), it will mean the production of approximately half a million eggs from one mature Benhamina each breeding season. This is more in agreement with the numbers produced by Nudibranchs than by Pulmonates.

Development proceeds at different rates in the same egg ribbon, those nearer the outside being further advanced than those in the inner capsules. It is thus possible to see trochopore, young and advanced veligers in the same egg ribbon. Later stages have not been observed, the entry of sea-water into the capsules during development causing death. Normally, disintegration of the ribbon begins about three weeks after it is laid, and this would free the more highly developed larvae on the outside first.

Lebour (1935), in describing the planktonic egg capsules of Littorina neritiodes. cites them as an exception to the general rule “for those molluscs inhabiting regions beyond or near high-water mark to be viviparous or have the veliger stage suppressed.” Benhamina also offers an exception to this rule, living at high-water mark, and having a well-developed veliger stage.


Members of the Siphonariidae living in the intertidal region present a particularly interesting problem. The possession of both gill and lung offers three evolutionary possibilities:


Siphonarias are in the process of leaving the marine environment for a terrestrial one. This would imply that the most advanced members of the group will live at the highest, tide levels.


Siphonarias having evolved so far, have ceased to progress. This also implies a gradient from low-tide levels occupied by more primitive types to higher tide levels with more advanced species.


Having achieved terrestrial existence, the group acquired a secondary gill and reverted to the marine environment. This postulates that the more primitive members will be found at the highest beach levels.

– 392 –

A gradient in habitat from low to high tide may be seen in New Zealand Siphonarias–S. australis living in the holdfasts of the kelp, Durvillea antarctica, at lowest tide mark, S. zealandica between low and mid-tide, Kerguelenella stewartiana above mid- and below high tide and Benhamina obliquata from mid- to high tide, sometimes raised above it by splash.

From anatomical studies, Hubendick (1945) contends that the primitive sub-genus of the Patelliformae is Liriola, from which the highly diversified sub-genus Siphonaria branched. Within the subgenus Liriola, two New Zealand species are classified—B. obliquata and K. stewartiana. Primitive features of Benhamina, e.g. cloven rachidian tooth, passage of genital ducts through part of the adductor muscle, wide opening to respiratory cavity, and general shell form ally it closely with the primitive S. lessoni group of Hubendick. S. cookiana and S. australis are classified in the sub-genus Siphonaria, of which S. cookiana is the most primitive member, and S. australis placed in the most advanced group, Siphonaria (sensu stricto). Thus anatomically. Benhamina is considered the most primitive of New Zealand Siphonarias, K. stewartiana, S. cookiana, and S. australis showing more advanced features in the order stated.

Correlated with range of habitat, this classification shows a gradient from the primitive Benhamina at, high tide to S. australis, which is the most specialised, at low tide. Therefore, it seems possible that the group, having achieved at least partial terrestrial existence, is returning to a marine environment. It would thus be expected that the gill, which by general agreement is stated to be of secondary origin, will be superseding the lung in more advanced forms. An investigation of this within the New Zealand species would be interesting.

Homing behaviour, which is an effective mechanism to resist desiccation and which is developed by molluscs of similar shell form living in the intertidal region, is exhibited only by certain individuals. Hence it is suggested that this is a recent development within the group, indicating that the period of time during which it has inhabited the upper tide levels is relatively short.


This paper represents a portion of a thesis presented for the degree of Master of Science in the University of New Zealand, and the writer wishes to thank Professor B. J. Marples, University of Otago, for advice about lines of attack; Miss Lucy Moore, Plant Researches Bureau, for identification of algae; and Mr. T. Aitken, Portobello Marine Station.



Mapping studies show that Benhamina obliquata occupies levels from mid- to high tide; Siphonaria zealandica from low to mid-tide in Otago Harbour.


B. obliquata is one of the dominant animals in the upper Bostrychia algal zone, and associated with the Lichina-Symploca associations; S. zealandica is dominant in the Hormosira-Corallina-Lichilla zones.

Picture icon

Cradle Rock Area, 176 sq. yds. Benhamina obliquata □, Siphonaria zealandica The contour lines each present one hour's rise in tide level from low tide (dark shading) to high tide (unshaded).

Picture icon

Area 400 yds. south-east of Portobello Marine Station. 194 sq. yds. Legend and contour lines as in Plate 41.

Picture icon

Area at Portobello Marine Station. 210 sq. yards. Mapped November, 1946. Legend and contour lines as in Plate 41.

Picture icon

Same area as in Plate 43. Mapped December, 1948. Legend and contour lines as in Plate 41.

– 393 –

B. obliquata is herbivorous, moving about only during the nocturnal low tide, activated by the stimulus of the ebbing tide.


Homing behaviour is used as the basis to divide Benhamina colonies into two ecological races of contrasting behaviour, termed Types A and B.


Age and rate of growth are estimated indirectly from the external winter rings of the shell.


The breeding season extends from late October until early March. Some aspects of courtship and egg laying are described.


Abe, X., 1935. On the Locomotion of a Limpet-like Pulmonate, Siphonaria atra Q. and G. Venus, Japan, vol. 5, (4), pp. 206–213.

Cottrell, A. J., 1910. Anatomy of Siphonaria obliquata Sowerby. Trans. N.Z., Inst., vol. 43, pp. 582–594.

— 1911. Vascular System of Siphonaria obliquata Sowerby. Trans. N.Z. Inst., vol. 44, pp. 374–379.

Crozier, W. J., 1921. Homing Behaviour of Chiton. American Naturalist, vol. 55, pp. 276–281.

Graham, A., and Fretter, V., 1947. Life History of Patina pellucida L. Jour. Marine Biol. Assoc. United Kingdom, vol. 20, no. 4, pp 590–601.

Hayashi, F., 1937. On the Locomotion Types of Certain Japanese Gastropods. Memoirs of College of Science, Kyoto Imperial University, (B), vol. 12. no 2, pp. 237–252.

Hubendick, B., 1945. Phylogenie und Tiergeographie der Siphonariidae zur kenntnis der Phylogenie in der Ordnung Basommatophora und des ursprungs der Pulmonatengruppe. Zoologiska Bidraa Fran Uppsala, vol. 24, pp. 1–216.

— 1946. Systematic Monograph of the Patelliformia. Svenska Vetenskapsakademiens Handlingar, Series 3, vol. 23, (5), pp. 1–93.

Hutton, F. W., 1882. On the New Zealand Siphonariidae. Trans. N.Z. Inst., vol. 15, pp. 141–145.

Lebour, M. V., 1935. The Breeding of Littorina neritiodes. Jour. Marine Biol. Assoc. United Kingdom, vol. 20, (2), pp. 373–378.

Mestayer, M. K., 1920. Notes on the Spawn Coils of Kerguelenia obliquata. New Zealand Jour. Sci. and Tech., vol. 3, no. 3, pp. 171–172.

Orton, J. H., 1926. On the Rate of Growth of Cardium edule. Part I, Experimental Observations. Jour. Marine Biol. Assoc., vol. 14, no. 2, pp. 239–279.

Risbec, J., 1936. Biologie et Ponte de Mollusques Gastédropodes Néo-Calédoniens. Bulletin de la Société Zoologique de France. vol. 60, pp. 387–417.

Russell, E. S., 1907. Environmental Studies on the Limpet. Proc. Zool. Soc. London, B., pp. 856–870.

Stephen, A. C., 1931. Biology of Certain Lamellibranchs on the Scottish Coast. Jour. Marine Biol. Assoc. United Kingdom, vol. 17, (2), pp. 277–300.

Vles, F., 1907. Sur les ondes pédieuses des Mollusques reptateurs. Comptes. Rendus Aca. Sci., Paris. Tome 145, pp. 276–278.

Willcox, M. A., 1905. Homing of Fissuralla and Siphonaria. Science, vol. 22, pp. 90–91.