Research Notes: Mobility, Gregariousness and Attachment in Four Small Bivalve Mollusc Species at Husvik, South Georgia

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

J. Moll. Stud.

(1995), 61,491^198 © The Malacological Society of London 1995

RESEARCH NOTES
Mobility, gregariousness and attachment in four small bivalve mollusc species at
Husvik, South Georgia
J. Davenport and P.C. Wilson
University Marine Biological Station, Millport, Isle of Cumbrae, Scotland, KA28 OEG, U.K.

Downloaded from http://mollus.oxfordjournals.org/ at University of Saskatchewan on September 19, 2012


South Georgian shores have four small bivalve currents that flow past the fronds of Macrocystis;
mollusc species living close to one another, between rough estimates with small floating pieces of wood
the sublittoral and the upper shore. Taxonomically indicate current velocities of 2-3 m. s"1. Kanin Point
they are disparate. Three are members of the Order is exposed to strong swell during periods of onshore
Heterodonta; Gaimardia trapesina trapesina winds. Gaimardia were obtained by collecting
(Lamark, 1819) (Family Gaimardiidae) and Kidderia Macrocystis with a grapnel and throwing line; all
bicolor (Martens, 1885) (Family Cyamiidae) are quite other species were collected intertidally. Specimens
closely related and have been placed in the same were transported to the Husvik Research Station,
family in the past, while their relationship with where they were held in sea water (changed daily) at
Lasaea rubra (Montagu, 1803) (Family Erycinidae) is 8-10°C. The salinity of the sea water ranged between
distant. Ussarca miliaris (Philippi, 1845) (Family 34 and 39%>. Animals were always studied within 4
Arcidae) is a member of the Order Taxodonta. days of capture.
Gaimardia (<10 mm shell length) attaches to the To assess gregariousness, 20 animals of each
giant subtidal kelp Macrocystis pyrifera (Linnaeus) C. species were removed from their holding vessels, all
Aghardh. Lissarca (<6 mm shell length) mainly lives byssus attachments being cut. The 20 specimens were
subtidally, but also occurs at low water spring tide spread out evenly on the floor of a petri dish filled
level. It is very common on intertidal red algae (par- with sea water and left for 24 hours at 8°C. After 24
ticularly Schizoseris condensata (Reinsch) R.W. h the number of animals that were still separate, and
Ricker) at the bottom of the shore, and also attaches the numbers attached to one another were counted.
to rock in surge channels. Kidderia is common on Additional qualitative observations were made on
(and under) lower shore rock and boulders. A slow- behaviour, using a binocular microscope.
gTowing species', it has a maximum length of 6 mm. Rate of movement was established for each species
High densities of Kidderia occur amongst the carpet at 8°C. To do this, animals were placed in a petri dish
of the lower shore alga Iridea cordata (Turner) Bory filled with sea water under which was placed a 1 mm
as well as on bare rock. The cosmopolitan Lasaea grid. The movement of animals was timed over a
occurs on the upper shore, either in crevices or (more distance of at least 3 mm. Results for 5 specimens of
commonly) on the underside of large rocks lying in each species were collected. Only straight-line, steady
pools. Small numbers of Lasaea also occur deep in movement was evaluated. The shell length of each
the Iridea carpet. Lasaea is the smallest of the four animal studies was also measured.
species, having a maximum length of about 3 mm. Specimens of each species were collected for study
All four species brood their young3, though breed- (by scanning electron microscopy, SEM) of byssus
ing strategies vary greatly otherwise, from the dio- form. Animals were taken directly from the environ-
ecious Lissarca to the self-fertilizing hermaphrodite ment, since subsequent byssus secretion in the labora-
Lasaea. Two of the species (Kidderia and Lasaea) tory might be atypical. The macroalgal dwellers
appeared to be gregarious, while the other two did (Gaimardia and Lissarca) were sampled by cutting
not (Gaimardia and Lissarca). All secrete a byssus, out pieces of seaweed that incorporated the byssus
but the form of the byssus attachment seemed to attachment. Kidderia were collected by scraping
differ considerably amongst the species. The present substantial numbers of them from the bare rock, so
study was designed to investigate gregarious be- that some individuals bound to others were collected
haviour, mobility and attachment, and to correlate without disturbance. Specimens of Lasaea rubra were
these features with the characteristics of the micro- collected in similar fashion, or were animals attached
habitats occupied by the four species. to small rock fragments lying in pools. All specimens
Animals were collected and studied at Husvik Har- were stored in 70% ethanol until returned to the
bour (54° ll'S; 36° 40'W) on the N.E. coast of South U.K., where they were dehydrated in an alcohol
Georgia during January-April 1994. Specimens of all series (70%-100%). After a final wash in dry 100%
species were collected at Kanin Point on the southern ethanol, the specimens were either critical point dried
shore of the fjordic harbour. Although the harbour is or air-dried in hexamethyldisilazane3, sputter coated
sheltered somewhat from onshore winds, fierce cata- with gold/palladium and viewed in a JEOL JSM 5200
batic winds often generate strong offshore surface scanning electron microscope. Additional obscr-
Downloaded from http://mollus.oxfordjournals.org/ at University of Saskatchewan on September 19, 2012
Figure 1. Byssus of Gaimardia trapesina. A. Single byssus thread emerging from the byssal aperture (arrowed).
Note that the thread has split during SEM preparation (scale bar = 100 u,m). B. Triangular adhesive pad
(attached to frond of Macrocystis). The byssus thread (arrowed) has broken off and the thin centre of the
triangle (t) has collapsed during SEM preparation (scale bar = 100 u,m).

Downloaded from http://mollus.oxfordjournals.org/ at University of Saskatchewan on September 19, 2012

Figure 2. Byssus of Lissarca miliaris. A. Multiple threads (arrowed) radiating from byssus opening onto
surface of piece of frond of Schizoseris (scale bar = 100 u,m). Bivalve (b) is at bottom of photograph, fragments
of alga (a) are at top. B. Closeup of slightly triangular attachment (c) of byssus thread (arrowed) to sea weed
(scale bar = 50 jun).
Downloaded from http://mollus.oxfordjournals.org/ at University of Saskatchewan on September 19, 2012
Kit
RESEARCH NOTES 495
Figure 3. Byssus of Lasaea rubra. A. View of two animals connected by byssus threads (scale bar = 500 u,m).
B. Closeup of byssus thread (arrowed) and attachment (a) to shell of conspecific (scale bar = 10 UJTI).

vations were made at Husvik under a binocular mobility largely ceased once a group had been
microscope. formed. Kidderia also formed numerous attachments
Equipment was not available to conduct a thorough to the substratum, as well as to conspecifics.
assessment4 of the mechanical properties of byssus. Table 3 shows speed of movement in the four
Instead, a crude estimate of byssus elasticity was species. The two species associated with seaweeds
obtained for 5 animals of each species in the following (Gaimardia and Lissarca) have similar relative speeds
manner. Each animal was left in a petri dish of sea- (shell length s"1) and move much more quickly (both
water for one hour so that it attached either to the absolutely and relatively) than the two species that
dish or to another animal. Beneath a binocular micro- attach to rock (Kidderia and Lasaea). Kidderia moves
scope, all byssus threads except one were cut. The much more slowly than does Lasaea, confirming the

Downloaded from http://mollus.oxfordjournals.org/ at University of Saskatchewan on September 19, 2012


'resting' undeformed length of the remaining byssus qualitative impressions gained in the gregariousness
thread (i.e. the length without any load being imposed) study.
was established (by eyepiece graticule accurate to 20 Gaimardia secretes few, usually single, long (often
M-m), then the animal was slowly pulled horizontally greater than the shell length) byssus threads (about
away from the byssus attachment until the byssus 60 (i.m diameter; Fig. 1A) that have a very strong
thread broke. The length of the deformed thread at triangular attachment to Macrocystis (Fig. IB);
the time of breakage ('failure' length) was established considerable force is needed to break the thread -
and the ratio between 'failure' and 'resting' lengths which always breaks before the attachment pad tears
calculated to give an ultimate extension ratio X5. All from the weed. The thread is clear and highly elastic
lengths measured were of the byssus thread external (Table 2), being capable of stretching about 5-fold
to the shell valves, so the short length of byssus within before breaking.
the shell was ignored. Lissarca also secretes highly elastic clear threads
All four species were observed extensively beneath (about 60 fim thick; Fig. 2) that attach to the sub-
the binocular microscope as part of this and other stratum by fairly strong, slightly triangular attach-
studies. In addition, during the period 28-31 March, ments. The threads are generally much shorter than
1994, there was an onshore storm with a sustained in Gaimardia, and the animal usually secretes 4-6
easterly swell. Four days later, the northern and threads that fan out from the byssal opening (Fig.
southern shores of Husvik Harbour were inspected to 2A). Sometimes, in weed, Lissarca also secretes a
assess the impact of the storms on populations of the multiple attachment thread along a piece of alga by
four bivalve species. sticking down the proximal end of the thread with a
From Table 1 it may be seen that Gaimardia and mass of byssus material, extruding a new length of
Lissarca are not gregarious in the sense of attaching thread, secreting another mass and so on. The effect
to one another. This was rather surprising in the case is to have 1 mm lengths of byssus thread 'tack welded'
of Lissarca as they occur in great densities on Schizo- to the alga. If the animal is pulled hard from such an
seris condensala at Husvik or on Desmarestia anceps attachment, the individual 'welds' may fail, yet leave
Montagne at Signy Island2. Microscopical obser- the animal attached to the weed by the next mass of
vation of dense clumps of Lissarca on Schizoseris byssus cement. The byssus thread of Lissarca needed
showed that few of the bivalves were attached to one less force to break than that of Gaimardia.
another; overwhelmingly they were attached to the Lasaea secreted several short, clear elastic threads
seaweed. (Fig. 3; Table 2) but these were extremely thin (< 12
In contrast, the other two species were highly u.m diameter), relatively weak and easy to break.
gregarious, forming large groups of animals bound Kidderia differed from the other species in several
together by a meshwork of byssus. There were, respects as far as byssus attachment was concerned.
however, behavioural differences between them. The byssus was thick (< 90 u,m diameter), soft and
Lasaea is extremely active and the animals, initially inelastic (Fig 4; Table 2), being incapable of sustain-
separated, were moving around their petri dish within ing even a doubling of length before failure. Kidderia
minutes. They did not lay down byssus on the sub- secretes several byssus threads and many threads
stratum, but moved around until they encountered a branch (Fig. 4A) so have multiple attachments.
conspecific. On doing so, one of the pair would attach Although individual threads are weak, much force is
itself to the shell (usually the umbone) of the other. needed to tear an individual animal from a bed of
Both would continue to move, though the direction Kidderia. The threads, through clear when secreted,
tended to be controlled by the larger animal. When rapidly become straw coloured and opaque.
the pair encountered other Lasaea, more attachment Gaimardia and Lissarca both have a large, strongly
took place and within 6 h all 20 of the Lasaea were adhesive foot. If their byssus attachment to a piece of
bound together in a single mass. However, the seaweed was cut, and the weed strongly agitated in
Lasaea remained quite mobile within the mass; indi- sea water, they were not dislodged, even though the
viduals appeared to find it easy to release their own shell was being swung violently and fro. In contrast,
byssus, or to break the byssus of others. Kidderia was the pedal adhesion of Lasaea and Kidderia was weak,
less active, took much longer to form groups, and the animals easily being dislodged from substrata by
Downloaded from http://mollus.oxfordjournals.org/ at University of Saskatchewan on September 19, 2012
RESEARCH NOTES 497
Figure 4. Byssus of Kidderia bicolor. A. Byssus mesh (arrowed; note branching) between two animals (b)
(scale bar = 100 u.m). B. Single adhesive disc (d) (scale bar = 50

Table 1. Gregariousness of four South Georgian bivalve species.

Species No. animals separate No. of animals attached to at


after 24 h. (Out of 20) least one other animal after
24 h. (Out of 20)

Caimardia trapesina 20 0
Lissarca miliaris 18 2
Lasaea rubra 0 20
Kidderia bicolor 0 20

Downloaded from http://mollus.oxfordjournals.org/ at University of Saskatchewan on September 19, 2012


Table 2. Elasticity of byssus thread in four S. Georgian bivalve species.

Species Ultimate extension SD Byssus thread


ratio M (n = 5) diameter (u.m)

Caimardia trapesina 4.97 1.67 60


Lissarca miliaris 3.85 0.94 60
Lasaea rubra 3.36 0.69 12
Kidderia bicolor 1.60 0.30 90

moving water if they were not attached by byssus. the weed. Second, recovery from dislodgement is
Gaimardia and Lissarca were both characterized by virtually impossible; Macrocystis, though attached to
an unusual reaction to touch. Most bivalve molluscs the substratum by a massive holdfast, is essentially a
(including Lasaea and Kidderia) respond to touch by tethered surface weed and a bivalve that falls from it
immediate foot retraction and shell valve adduction. is most unlikely to be able to regain contact with kelp.
The two weed-dwellers barely responded to a light A highly streamlined shell and strong, elastic byssus
touch, while stronger tactile stimulation provoked counteract these problems. Because the byssus thread
brief shell valve adduction, but not foot retraction. is very long, the animal can move rapidly around on
Only prolonged stimulation could invoke a 'normal' the weed without losing a firm attachment. During
closure response and then only for a few seconds. periods when the animal detaches the byssus the
Garimardia and Lissarca both secreted byssus very unusually strong pedal adhesion provides further pro-
rapidly (often in less than 60s) when taken from sea- tection. The overriding necessity of avoiding dis-
weed and placed in holding vessels. Attachment time lodgement presumably explains the suppression of
in the other two species was very variable and much the normal bivalve closure response to touch.
slower; several minutes often elapsed before byssus Lissarca faces similar problems, particularly when
secretion. living on intertidal Schizoseris. This complex red alga
Observations in early April confirmed the pre- is swept to and fro on every wave at the time of low
carious nature of the lifestyle of Gaimardia and water, presumably causing oscillations in inertial
Lissarca. After the storm of 28-31st March, 1994, forces acting upon the Lissarca. A highly elastic
large numbers of both species were found either dead byssus will help to absorb these shocks. Lissarca often
on the shore or living on the 'wrong' substrata. Several occurs on bare rock around Schizoseris plants, es-
Gaimardia were seen attached to middle shore pecially in surge channels. It seems probable that
Porphyra or lower shore Palmaria, specimens of Lissarca are sometimes dislodged from Schizoseris —
Lissarca were also seen on Porphyra, or on bare rock, they may well be able to regain their position by
but there were many floating in the surf, their shells virtue of a strongly adhesive foot and high level of
filled with air. In contrast, there was no evidence of mobility. The suppression of the normal shell valve
dead or dying Kidderia or Lasaea. closure response in both Gaimardia and Lissarca is an
The four species show a diversity of approaches to interesting example of convergence, given their
solving the problem of attachment in a high energy taxonomic separation.
environment. Gaimardia, living on Macrocyslis, faces Kidderia often lives on bare rock, battered by
two main problems. First, the strong wind-generated waves. It is unlikely to encounter the inertial prob-
surface currents and exposure to oceanic swell mean lems associated with being swept to and fro in weed
that the animals are exposed to considerable and - when living amongst Iridea carpet it lives amongst
highly variable forces tending to dislodge them from the matted holdfasts, not the fronds. Kidderia clearly
498 RESEARCH NOTES
does not need highly elastic byssus; instead it requires
1
a tightly bound, firm attachment; it also relies on
gregarious attachment, often with thousands of ani-
in JZ
II
O O> <D ^ mals bound together in a soft meshwork of thick,
oO «- Q O O

(shell len
C II
c p S q p branched fibres. The meshwork also provides a safe
dodo microhabitat for newly-released brooded young.
Rapid movement is unnecessary in this species.
Lasaea rubra lives in a more protected environ-
ment, avoids exposed rock or plant surfaces. The
lower risk of dislodgement permits lower pedal ad-
hesion, a relatively weak and fine byssus, yet a high
7 level of mobility (allowing invasion of new microhabi-
"o
13 c
CO tats).
CD 0 j j One of the authors (JD) acknowledges the financial

Downloaded from http://mollus.oxfordjournals.org/ at University of Saskatchewan on September 19, 2012


CD
Q. O)Ec support of the Royal Society and of the TransAntarc-
CO
> _o tic Association, plus the logistic backup of the British
c o
CO Antarctic Survey (BAS), which enabled him to work
CD
"5
(0
on South Georgia. Thanks are due to Dr D.H.
Walton and Dr W. Block of BAS who encouraged
him to work at Husvik.

i REFERENCES
o
in *~ i- to i- .-
Q II X
i- p p p
1. Ralph, R. & Everson, I. 1972. Brit. Ant. Surv.
CO II Bull., 31: 51-54.
1 d d d d
— CO
2. RICHARDSON, M.G. 1979. Brit. AM. Surv. Bull.,
E 49: 91-115.
3. COOLEY, W.A., SCOTT, A.C., & SIMMONS, M.M.
1994. Microsc. & Analysis, 43: 23-25.
4. SMEATHERS, J.F. & VINCENT, J.F.V. 1979. J. Moll.
Stud., 45: 219-230.
I
c
CD O 5. DENNY, M.W. 1988. Biology and the mechanics of
CD *~
00 Q.
i^ in co co the wave-swept environment. Princeton University
CO CD X i- p p
Jo C O •-
CNI
o o d ci
Press, Princeton.
ies

CO 1
CD CO
^
Q.
*B —
CO
CD

"5
gian bi\

in
(mm)

Q
CO
II co in «- <6
c t^ d ^ d
o
CD

CO _
^_ CD
3 ^
CO CM CM
c c p in
CO

c
CD 3 " CM co

CD
E
CD
O
E
"o 5
CO CO

CD
Q. §|58
III?
CO
CO 0)
CD
CD "o
n
£
CO ijil

You might also like