MARINE ECOLOGY PROGRESS SERIES

Mar. Ecol. Prog. Ser.

Vol. 86: 217-227,1992

Published September 24

Dispersal of echinoderm larvae in a geographical

area marked by upwelling (Ligurian Sea,
NW Mediterranean)
Maria Luiza Pedrotti, Lucienne Fenaux
Laboratoire d'Ecologie du Plancton, Universitb P. et M. Currie, Observatoire Oceanographique de Villefranche-sur-Mer.
La Darse, F-06230 VillefrancheIMer, France

ABSTRACT: A spatio-temporal survey of some echinoderm larvae was conducted along a transect of
28 nautical miles from the Bay of Villefranche, France, halfway to Corsica. Distribution patterns and
mechanisms responsible for larval dispersal, viz. water currents, hydroclimate and vertical movements,
were studied. The echinoderm larvae exhibited weak vertlcal migration (a few meters), remaining in
the surface layer. They are therefore subjected to hydrodynamic constraints which either limit or promote dispersal away from the nursery beds. Larval concentration decreased with distance from the
coast and was generally restricted to within 16 miles of the coast This distance coincided with the
location of a coastal divergence zone where rising subsurface waters confined larvae to the Ligurian
Current.

INTRODUCTION
At the beginning of the pelagic phase, the horizontal distribution of marine invertebrate larvae corresponds to that of the parents and thus to favorable
habitats for settlement (Young & Chia 1987). Currents
and turbulence cause larval transport and dispersal
(Okubo 1980). The transport of marine invertebrate
larvae by currents has been well studied by Scheltema (1966, 1968, 1971, 1972, 1975, 1986) and Mileikovsky (1966, 1968, 1970, 1974). The larvae may be
dispersed and recruited to other distant communities
(Thorson 1961, Mileikovsky 1971, Scheltema 1974).
For meroplanktonic animals with pelagic planktotrophic larvae, the amount of geographical dispersal
increases with the length of larval life and depends
upon such factors as swinuning behavior and hydrodynamics. Whereas biological interactions appear to
be responsible for the horizontal distribution of
oceanic zooplankton (Haury 1976, Haury & Wiebe
1982), physical processes are the principal agents in
coastal regions (Haury et al. 1978, Greenblatt 1982,
Yamamoto & Nishizawa 1986). Thus, the spatiotemporal distribution of zooplankton populations, on
the medium scale, depends on the movement of water
masses, upwelling, convergences and vertical mixing
O Inter-Research 1992

(Mackas et al. 1985, Boucher et al. 1987, Ebert & Russell 1988).
In the C6te d'Azur region (French Riviera), the narrow continental shelf limits the bathymetric distribution of neritic echinoderms and consequently restricts
the source of planktotrophic larvae strictly to the coast.
Some larval production remains near the coast, thereby augmenting its chances of finding nutritive zones
for larval growth and settlement, while the remainder
is transported by currents into open waters. For neritic
invertebrates the major problem is reaching nursery
zones conducive to larval settlement and juvenile
recruitment.
Under what conditions does larval production result
in success? More specifically: What are the mechanisms which retain larvae in coastal environments?
Are there ecological barriers which limit larval dispersion towards open waters? Are larvae passive or active
in regard to hydrodynamics?
The objective of the present study was to describe
the seasonal and temporal variations in the distribution
and abundance of echinoderm larval populatlons and
to relate this to processes which determine larval dispersal, viz, currents, temperature and salinity distribution, and vertical and horizontal water movements.
Plankton tows were made in the Bay of Vlllefranche-

Mar. Ecol. Prog. Ser. 86: 217-227, 1992

218

sur-Mer and also along a 28 nautical mile transect perpendicular to the coast between Villefranche and Calvi
(Corsica).

&

Peripheral Frontal
L

;;.
5

0"

Central

.
L
H

E (
3

STUDY AREA

The study area was in the Liguro-Provenqal Basin in

the northeast part of the western Mediterranean. This
basin is the center of a cyclonic circulation (Bethoux et
al. 1981, Bethoux & Prieur 1983) induced by largescale integration of climatic factors (evaporation and
precipitation) into the general circulation of the
Mediterranean Sea (Prieur 1983). Surface waters from
the Tyrrenean Sea and the west coast of Corsica form
the Liguro-Proven~alCurrent. From the Gulf of Genoa,
this current flows westward along the Italian and
French coasts. The Ligurian Current occupies a zone
stretching from the coast to between 15 and 20 miles
offshore; its width varies according to season (Bethoux
et al. 1988). The average current flow is on the order of
1.4 X 106m3 S-' and the average speed in surface layer
is 20 cm S-' between 3 and 8 miles offshore and 30 cm
S - ' between 8 and 14 miles offshore (Fig. 1).

35 miles

pp~qrC

zd!.

.-

..g.......
2 .............-**

Fig. 2. Schematic representation of the circulation along the

transect studied showing the peripheral, frontal and central
zones, and convergences (C) and divergences (D). (After
Boucher et al. 1987)

and 0.2 to 0.4 kg m-3in winter (Boucher et al. 1987).In

the vertical, this front can be recognized by sloping of
isopycnal lines and by an active vertical circulation
with alternating convergences and divergences
(Fig. 2). The hydrological structures of the Ligurian
front are permanent; the front is very visible in the surface in winter and early spring, but varies in intensity
throughout the year.
Bay of Villefranche. This bay is situated on the
northern side of the Ligurian Sea and measures 3.5 km
along its north-south axis (Fig. 3). The 150 m isobath
skirts the entrance to the bay. The depth decreases
northward to a sill of 15 m. Surface water masses are
regulated by 2 processes: vertical stabilization and
homogenization. The bay is subject to the influence of
a cyclonic current running from east to west along the

Fig. 1. Schematic map of the circulation in the LiguroProvenqal Basin. The horizontal transect studied is indicated

Hydrological front. Owing to the cyclonj c circulation, the surface waters are separated into a central
dense homogeneous zone and a less dense peripheral
zone. Between these 2 zones a density front 1s formed
m.arked by very pronounced horizontal density and
salinity gradients (Sournia et al. 1990).The mean horizontal density gradient in the frontal zone is higher
than in the other zones; the horizontal difference in
density near the surface is 0.6 to 1 kg m-3 in summer

150 m43'40 N
Fig. 3. Locations in the Bay of Villefranche where plankton
were collected (Stn S) and salinity and temperature were
measured (Stn B)

Pedrotti & Fenaux: Dispe:rsal of echinoderm larvae

CGte d'Azur. Under normal conditions, a branch of the

Ligurian Current flowing westward offshore enters the
surface of the bay and spreads out northward. Where
the depth is greater than 30 m, the current sinks and
leaves in an opposite direction to the entry current;
where the depth is less, a surface contrary flow is
created. Impulses related to the west wind (Mistral)
tend to reverse the circulation, pushing the surface
mass offshore and replacing it with subsurface water.
In the middle part of the Bay the flow is 5 cm S-', and it
decreases to 2 cm S-' in the northern part (Hentsch
1959, Nival et al. 1975) (Fig. 4). The thermal stratificaBay of Villefranche
North

South

(z)

Fig. 4. Vertical section of the Bay of Villefranche showing the

circulation in weak wind conditions. A surface current enters
the bay and spreads out northward. If depth is greater than
30 m the current inverses and leaves in a n opposite direction
to the entry current; if depth is less than 30 m, a surface contrary flow is created. (After Nival et al. 1975)

tion allows nutrients, which enter via urban effluents,

to enrich the waters which tend to stagnate in sheltered areas of the bay. This enrichment favors phytoplankton development (Bethoux & Prieur 1983).
General description of annual variations in temperature and salinity along the Villefranche coast. In the
coastal region of Villefranche temperature and salinity
data have been collected weekly since 1957 (Stn B;
Fig. 3). Between January 1984 and December 1986 at
10, 30 and 75 m depth temperatures ranged from 12 to
24 "C and salinity from 37.2 to 38.4 %o. The temperature is homogeneous in winter throughout the water
column. From the end of March, it rises progressively,
reaching a maximum of 24 "C in August. The water
column is stratified until the end of summer. The temperature then decreases until February of the next
year (13.37 "C on the coast, and 13.5OC at 28 miles offshore). In winter, under the influence of a cold, dry
continental air mass, the density of the surface water
increases, generating vertical mixing of the water column. In February, a zone of high salinity appears between 13 and 18 miles offshore and most likely indicates the leveling off of the intermediate Levantine
layer (Gascard 1991), generally situated at 400 m
depth near the coast and 200 m offshore, and the for-

219

mation of the winter front in the surface. From the end

of May, the surface layer begins to stratify. Beginning
in August, the 0 to 30 m layer becomes colder and
denser, and there is a tendency towards the formation
of a homogeneous layer.

MATERIAL AND METHODS

Spatio-temporal surveys of echinoderm larvae

began with the analysis of samples collected by the
research vessel 'Korotneff' between March 1984 and
February 1985, and during 1986 along the 28 mile
transect from Stn B at Villefranche halfway to Calvi
(Corsica). To these series plankton tows were added,
starting in September 1986, at Stn S in the Bay of
Villefranche.
Plankton tows. At all stations, tows were made with
a 200 pm mesh net. In the bay, 5 vertical tows (30 to
0 m) were made each week. During 1984 and 1985,
vertical tows (200 to 0 m) were taken along the transect twice a month at 6 stations, 5 miles apart. In 1986,
horizontal tows were made with a high speed Hai
sampler (Boucher 1984) modified by Fenaux & Pedrotti (1988). Plankton was collected continuously at
10 m depth. An inboard peristaltic pump brought the
animals from the net collector to 2 sampling schemes
alternatively. Each sample corresponded to 5 min of
pumping with an average ship speed of 5 knots (i.e. a
horizontal distance of about 750 m). Between February and June tows were made from the Bay of Villefranche to 11 miles offshore (about 30 samples were
preserved during each trip); between July and December samples were taken from the Bay to 28 miles
out (about 70 samples were made during each trip).
The depth was chosen after statistical analysis of
vertical tows showed that abundant larvae were to b e
found in the first 10 m (Pedrotti 1990).
Salinity, temperature, and pressure measurements
were made with a 8705 Guildline CTD probe from
800 m to the surface (PROS VI program; Prieur et al.
1990).During zooplankton collection with a pump net,
the temperature was also constantly measured using a
probe attached directly to the sampler at 10 m depth.
Salinity samples were also collected every 750 m and
measured using a conductivity salinometer (Autosal
8400). Densities were calculated, using the Unesco formula, from temperature and salinity and presented as
density excess, which shows the excess density of sea
water compared to 1000 kg m-3 water.
The abundance of planktonic larvae is expressed as
the number of individuals collected per 10 m3. The
different volumes filtered by the net were calculated,
using equations from Wieghardt (1953) and Tranter
(1967). The tows at Stn S in the Villefranche Bay (30 to

Mar. Ecol. Prog. Ser. 86: 217-227. 1992

220

0 m) were 7 m3; the vertical tows (200 to 0 m) were

48 m3. We estimated the volume of filtered sea water in
the continuous tows using one flowmeter (Hydro-bios
model 438110) situated at the center of the net and
another outside the net. For each 750 m horizontal
transect, the volume of sea water filtered was 173 m3.
After each collection, the plankton was immediately
fixed in 4 % formaldehyde. We analyzed the samples
from the stations in their entirety using a stereoscopic
microscope. For this study 2 classes of larvae were isolated and counted: echinoplutei and ophioplutei. Two
ophioplutei abundant in our samples, Ophiopluteus
bimaculatus (Miiller) and 0. compressus (Mortensen),
could not be assigned to species; these names correspond to the larval names given by MiiLIer (1852) and
Mortensen (1927).

m
0
d
U
.d

VI

RESULTS

Distribution of echinoderm larvae in the LiguroProvencal Basin

The area sampled is characterized by the presence of
a late-spring population of echinoderm larvae and by
an early-autumn/winter population in which the larvae
are more abundant. The spatial distribution of echinoderm larvae is in general marked by a decrease in the
number of larval species with increasing distance from
the coast, with an outer limit of distribution within
13 miles of the coast (Figs. 5 & 6). However, at the end
of November and December 1984 larval abundance at
offshore stations was greater than at stations nearer
the coast (Fig. 7).
In samples taken in 1984 the lowest concentration of
echinoderm larvae were observed in March and A p d
and coincided with a period of abundance of Salpa
fusiformis in the region (Fig. 7). In May the increase in

y
0

400

-22
-2

I Average

I Maximum

220 -

: D,L

300-

-A&&

,-

13

18

28

miles

Fig. 5 Echinoderm larval abundance in samples taken along

a transect from the coast to 28 miles offshore (6 stations).Grey
bars: maximum larval ab.undance found at each stat~onbetween March 1984 and February 1985. Black bars: average
larval abundance for each station during 15 cruises along the
transect studied

t-

01 5@nP!,'!P"l

g
d
ti

aJ

221

Pedrotti & Fenaux: Dispersal of echlnoderm larvae

29 March 1984

09 April 1984

M ! .

.r.3

13

18 24

28

24 May 1984

150

400 -

18 24

28

13

18 24

28m1es

02 July 1984

20 September 1984

16 July 1984

Fig. 7. Number of echinoplutei a n d

ophioplutei In samples between
March 1984 and February 1985,
from the coast (3 miles) to 28 miles
along the transect. Samples from
7 May a r e omitted because there
were no individuals. Note that vertical axes a r e not to the same scale

13

I9 June 1984

13

18 24 28

13

18 24

28nul~~

16 October 1984

19 November 1984

03 December 1984

14 December 1984

2 1 January 1985

04 February 1985

larval abundance corresponded to the presence of

young Amphiura f W o m i s larvae and to the spawning
season of Paracentrotus lividus and Arbacia liwula. In
June and July larvae become older and competent to
n~etamorphosis.The absence of younger echinoplutei,
which were abundant during the preceding month,
showed that this is the end of the reproductive period
for several echinoderm species. In November, very
young larvae of A , filiformis, Ophiopluteus bimaculatus and Ophiopluteus compressus were found nearshore (3 and 8 miles, respectively) while older and
competent larval stages were collected at stations far
from the coast (23 and 28 miles). In early December
post-larvae of A. filiformis were found at stations 18
and 28 miles offshore.
Data from horizontal continuous tows during February to April 1986 showed a paucity of plankton
(Fig. 6). At the end of May spring populations of Paracentrotus Lividus and Arbacia lixula appeared as in
1984. In June, more than 80 % of echinoplutei of P
lividus had developed a rudiment. In July, A. h u l a
echinoplutei with 8 to 10 arms, several spatangoids,

13 18 24 28

13

l 8 24

28mlleS

and O p h o t h n x fragilis and Ophiopluteus bimaculatus

larvae were abundant to 10 rmles out. At the end
of September and beginning of October, the postmetamorphc stages of A. Lixula were hspersed between ? and l ? miles. From the beginning of October
to December, ophioplutei were more abundant than
echinoplutei in pump samples along the transect.
Young echinoplutei of P lividus and A. lixula appeared at the end of October. Up to this tinle, it was
possible to follow the development of a larval cohort
of P lividus. Early November larvae were at the 8-arm
stage, and were then replaced by older stages at the
end of November until the post-larval stage. Amphiura fiLifomislarvae, Ophiopluteus compressus and 0.
bimaculatus appeared from the end of October to the
end of November. In December larvae disappeared
(detailed information on spatial distribution of different larval stages of echnoderm larvae will be published elsewhere). Whatever larval stages or species,
there was a drastic decrease in echinoderm larvae
beyond 16 rmles in continuous tows during October
and November.

Mar. Ecol. Prog. Ser.

222

Biological and physical events possibly related to

larval distribution
Food supply and larval competition

PKOS \ I C l .rTIO\

0
I

n
:

4 1

80 M)

To understand the mechanisms controlling larval

distribution in the Ligurian Sea, several factors must be
taken into account. Firstly, the temporal distribution of
larval populations depends on their biological cycle,
and on their functional adaptation and teleological
strategies. Food supplies are distributed in the sea according to geographical area and seasonal changes.
For this reason, it is complex and difficult to determine
the mechanisms by which nutritional resources affect
larval cycles in the field. In the Liguro-Provencal Basin,
plutei find oligotrophic waters. Fenaux (unpubl.) observed that the development of a planktonic cohort of
Paracentrotus lividus, followed during a phytoplankton bloom in the Bay of Villefranche in 1986, is longer
than the development of larvae reared with a n enriched diet. Perhaps larvae become adapted to poor
environmental conditions by increasing the length of
larval life. An extended planktotrophic period may increase the mortality rate by predation and the potential
of dispersal and reduce chances for the larvae to return
to the coastal areas of Villefranche.
The low larval concentration observed during March
and April may be more related to biological interaction
than to larval dispersal. Samples taken daily at Stn B
in the Bay of Villefranche showed the presence of larvae of the Spatangoidea Echinocardium flavescens
(Miiller),Ophiopluteus compressus and Amphiura filiformis in early March 1984 (Pedrotti unpubl.). The
drop in larval abundance in the middle of March at
Stn B and on the transect coincided with the appearance of a bloom of Salpa Eusiformis (Tunicata) in the
region (Choe 1985). Due to the reproductive adaptations of salps, namely high reproduction rate (forming
dense blooms) and high individual rates of growth,
they can supplant other filter-feeding species (Braconnot et al. 1988).

Hydrological events
Physical parameters were measured during the same
period that zooplankton were collected. In November
1.984,the distribution of larvae up to 28 miles offshore
was related to the physical parameters. At this time,
salinity and temperature in surface waters to 50 m
depth were nearly homogeneous from the coast to
28 miles offshore; the density front was present between 15 and 23 miles, but below 50 m. (Fig. 8 ) .Probably larvae could be transported offshore due to the
absence of a density frontal barrier in the surface layer

,
,

:i

,..:..,..
,

-.140 L

160 1811 -

TEMPERATURE

Fig. 8. Temperature and salin~tyon 19 November 1984, from

the coast to 28 miles along the transect. Data are from the
PROS V1 programme

During autumn 1986, high salinity occurred along

the coast, up to between 14 and 17 miles offshore. A
similar phenomena was also observed in 1957 and
1958 by Bougis & Carre (1960) and in 1987 and 1988
by Dallot & Pizay (1989). The following scenario may
explain the succession of events. During the summer,
the surface of the Liguro-Provencal Basin is the site of
transformation of Atlantic waters into those of the
Mediterranean which are warmer and more saline
(Gostan 1968). The current which enters the region in
autumn, coming from the south, has a higher temperature than the waters in the basin. The lack of rain
during summer 1986 resulted in an increase in salinity
of surface waters. The high evaporation rate without a
decrease in air temperature resulted in a positive
thermal balance. Very weak winds allowed a stable
and homogeneous surface layer. The density increases from the coast (peripheral zone) to 28 m ~ l e s
out agree with the conceptual scheme established for
this region.
The limits of larval distribution during October and
November 1986 corresponded with changes in physical parameters. At the end of October, comparing
stations on either side of 16 miles offshore, the number
of larvae per 10 m3 dropped from 30 to 5 individuals
(Fig. 6); further away no larvae were found. The surface tempera.ture decreased from 19.80 to 18.00"C and
salinity from 38.27 to 38 1 3 ' h (Fig. 9). In November,
the abundance of larvae at stations around 14 miles
offshore decreased from 35 ind, per 10 m3 to 10 ind.
(Fig. 6 ) . Beyond 14 miles offshore no larvae were
found. The surface water temperature decreased from
17.50 to 16.30 "C and salinity from 38.30 to 38.14 X".
The differences In density excess near the s.urface
either side of 16 miles offshore in October and of
14 miles offshore in November ~ 7 e r e0.29 k g m - 3 and
respectively. These differences corre0.30 kg m
spond to the density gradient across the front.

"

223

Pedrotti & Fenaux: Dispersal of echinoderm larvae

30 October 1986

06 November 1986

B
ot

28.50
28.00
27.75
27.5 0
27.25
27.00
0

14

27.00
28 miles
0

21

14

21

28 miles

Fig. 9. (A) Salinity a n d temperature measured at 10 m depth on 30 October a n d 6 November 1986, from the coast to 28 miles along
the transect (B) Density excess (sigmat-t) in k g m-' calculated from the temperature a n d salinity given in the upper graphs

Distribution of Paracentrotus lividus larvae in the

Bay of Villefranche

The succession of the different larval stages of

Paracentrotushv~dusat Stn S in the Bay of Villefranche
from September to December 1986 was the same as
that observed along the transect. Larvae with 4 arms
appeared at the end of October (Fig. 10).Larvae with 8
arms appeared in early November; the rudiment stage
was collected between 7 and 19 November and postlarval stages at the end of November. Undoubtedly,
advanced stages and post-larvae found in the Bay belonged to the same population. The same succession of
different larval stages as found in autumn 1986 was
found in 1987 (Pedrotti 1990). These results suggest

,,,_

Paracenrrotus lividus

September

October

Station S

November

December

Fig. 10 Paracentrotus l i v ~ d u s Different larval stages (ind

IOrn-' in the Bay of Villefranche (Stn S ) from September to
December l986

that the larvae sampled come from a simultaneous

autumn spawning that favours synchronous larval development. A part of the larvae produced by the local
and nearby adult populations were retained in the bay
by hydrological processes. During autumn 1986, after a
period at the end of October during which strong
winds seem to have stimulated gamete release
(Pedrotti unpubl.), there was a period in which winds
became weaker. In November, easterly winds were
predonlinant. In such conditions, the surface current
entering the Bay has a reduced flow speed and the larvae can probably remain for several days trapped in
the circulation by the contrary flow or by local eddies
created in the surface layer (see Fig. 4 ) .
Phytoplankton development, d u e to enrichment of
nutritive salts, can also be a factor favoring larval d e velopment in the Bay. The Paracentrotus lividus larval
development observed at the end of May 1986 in the
bay coincided with the e n d of the spring phytoplankton bloom (Fenaux et al. 1987). A series of continuous
plankton tows made parallel to the coast between
Antibes and Monaco (French Riviera) in J u n e 1987
showed the presence of larval peaks in relation to
coastal features; there were higher densities off the
Bay of Angels (Nice) and off the Bay of Villefranche
(Pedrotti 1990).
While hypotheses concerning physical mechanisms
of larval retention can be supported by our results, predictions about how long a cohort of larvae must persist
in the Bay to allow recruitment success are uncertain.
In attempting to estimate the decreasing abundance
of Paracentrotus lividus larvae with stage the number
of individuals in a cohort, from the appearance of 4-

Mar. Ecol. Prog. Ser.

224

Table 1. Paracentrotus lividus. Decrease m abundance of

larvae from the day of the appearance of the 4-arm stage
(100 %) to the planktonic postlarval stage (PL) at Stn S in the
Bay of Villefranche. Boldface type shows highest percentage
of each stage. Rud: 8-arm stage + rudiment
Day

1
2
3
4

5
8
9
10
11
12
16
l?
18
21
22
23

% 4-arm
100

49.7
3.0
1.2
0
0.61
0
0
0
0
0
0
0
0.61
0.61
0

% 6-arm

O/o %-arm

% Rud

% PL

0
4.2

0
0.6
0.61
1.8
2.4
7.9
13.9
2.4
1.2
0
0
0.61
0
0
0.61
0

0
0
0
0
0
0
1.8
3.0
3.0
1.8
7.9
8.5
2.4
2.4
3.0
0

0
0
0
0
0
0
0
0
0
0
0
0
0
0.6

13.3

1.8
0
0.61
0
0
0
0
0
0
0
0
0
0

1.2
1.2

arm stages to the appearance of post-larval stages, was

followed regularly at Stn S in the Bay of Villefranche.
Taking the 4-arm stages as 100 %, only 1.2 % postlarval stages were observed after 23 d (Table 1). If one
considers that the larvae sampled come from the same
spawning originating in local coastal Villefranche population~and that a part of them are retained in the bay
by the bioloqcal and physical mechanisms described
above, the local density of older larval stages corresponds closely to recruitment. The decrease probably
corresponds to mortality and dispersal. The decrease
in larval abundance is often falsely attributed principally to mortality, as pointed out by Strathrnann (1985).
Even if the mortality rate in the field is considered to be
very high (Vance 1973a, b, Chia 1974), the role of
dispersal in population dilution is also important. The
estimation of the mortality rate becomes valid if it is
made on populations having synchronous gamete
batches and where advective and diffusion transport
can be estimated (Korringa 1952).

DISCUSSION

Larval distribution perpendicular to the coast

Boucher et al. (1987) studied the space-time microdistribution of zooplankton in relation to hydrological
phenomena in the Ligurian Sea. T h s first study classed
echinoderm larvae as a group belonging strictly to
coastal areas and corresponding to the maxima of
spring biomass. The frontal zone was considered as the

offshore boundary which is not crossed by neritic

species. Our results show, however, that larvae and
post-larvae display a great potential for dispersal and
their distnbution is directly associated with variations
in regional hydrological phenomena. The larvae and
post-larvae of echinoderms, which have no ontogenetic migration, a die1 vertical migration of a few
meters and a slow swimming rate (Banse 1964, 1986,
Pennington & Emlet 1986, Pedrotti 1990), may act like
passive particles subjected to hydrodynamic controls
associated with surface waters. Currents sometimes
represent barriers to dispersal; sometimes they may
enhance larval dispersal (Scheltema 1988, Scheltema
& Rice 1990), transporting them away from favorable
sites for settlement and juvenile growth.
In the studied region, at the end of November 1984,
when the extension of the Ligurian Current is maximal, larvae were observed as far out as 28 miles, at the
transect station furthest from Villefranche. Two hypotheses could explain larval presence at 28 miles: (1) the
front does not manifest itself at the surface during t h s
period, and no barrier exists to limit the offshore
extent of larval dispersion (2) during autumn there is a
gradual transition between the coastal system
(Ligurian Current) and the system in the open sea
(Bong 1983) due probably to eddies resulting from instabilities in the surface layer of the current, where
echinoderms are present (Louis Prieur pers, comm.). In
the absence of discontinuities of physical parameters
larvae are transported by the mainstream of the current to the open sea. Hence, the numbers of individuals recruited in the areas where adults are found will
be reduced.
In October and beginning of November of 1986, during the spawning autumn season, the limits of larval
distribution corresponded with a hydrodynamic barrier. The presence of a coastal divergence zone, where
there is an upwelling of subsurface waters, most likely
limits the larval distribution in the Ligurian Current. A
schematic drawing showing horizontal larval distribution and hydrological conditions during autumn 1986 is
given in Fig. 11: larvae are distributed up to 16 miles
out. This distance coincides with the edge of the
Ligurian Current. The stratification of the water column keeps larval distribution near the surface waters
in autumn.
Contrary phenomena were observed by Pedrotti
(1990) during winter 1986. At this time, due to the
vertical mixing of the water column, larvae are passively transported from the surface layers downward
according to patterns of circulation. Larvae that are
weak swimmers and that do not migrate vertically are
incapable of traversing discontinuities, and hydrodynamics alone can explain the presence of the larvae
at depth.

225

Pedrotti & Fenaux: Dispersal of echinoderm larvae

Recruitment chances for larvae found further

offshore
How do larvae return to the coastal regions after
being dispersed to open waters ?
The most important mode of transport shoreward of
the megalopae of the crab Pachygrapsus crassipes is
via surface slicks associated with tidal internal waves
(Shanks 1985). This transport is also the means by
which some other benthic invertebrate pelagic larvae
return to the coast (Shanks 1983, 1986, Shanks &
Wright 1987). In the Ligurian Sea tides are negligible
and the bottom depths do not intervene in the circulation of the water masses; it is unlikely that larvae could
be trans~ortedshoreward bv such a mechanism. A
wind-driven surface water mechanism, as found by
Emlet (1986) in East Sound (Washington, USA) to
transport advanced larval stages of Dendraster excentncus to adult habitats, is equally unlikely. In the studied area, the direction and intensity of the wind contribute to the circulation near the coast (in the Bay of
Villefranche), but play little part in the cyclonic circulation of the Basin.
Besides favorable conditions for larval retention observed in the Bay of Villefranche we cannot hypothesize about the development of echinoderm populations
found near the coast due to the presence of the
Ligurian Current. Along the coast, the Ligurian
Current has a complicated circulation, forming meanders with possible upwelling near the coast due to secondary circulation (Sournia et al. 1990; Fig. 2). Those
larvae found in the vortices will have better chances of
being transported to nursery zones, probably in bays.
In comparison, plutei found in the main flux of the
Ligurian Current (2000 m depth) or transported to the
open sea Mrlll probably not return to the same coastal

Peripheral zone (autumn 1986)

frontal

dge
of ,he

zone

Ligurirn Cumnt

Villefranche
f 17 miles

1 ----I----

Fig. 11. Vertical section from the coast to offshore, showing

larval distribution and hydrological conditions during autumn
1986. Larvae are distributed up to 16 miles out. This distance
coincides with the coastal hvergence zone, where there is a n
upwelling of subsurface waters, and this most likely limits the
larval distribution in the Ligurian Current. The density excesses (sigma-t) in kq m-3 were obtained from the PROS V1
( ~ r i e ueti al. 1990) programme

Liguian Current

Fig. 12. Schema showing the possible trajectory of larvae in

the Ligurian Current during autumn 1986. Along the coast,
meanders are formed, whereas in the open sea, the current is
much stronger. Larvae found in the meanders will have the
best recruitment chances

regions. They must prolong their larval period until

finding favorable recruitment sites (Fig. 12). The presence in November 1986 of larvae further out which are
older than those found on the coast suggests that larvae delay metamorphosis. This problem is posed for
competent larvae of the infra-littoral species Arbacia
h u l a and Paracentrotus lividus, for which the greatest
depths of distribution in the Mediterranean sea are 40
and 80 m, respectively. In the laboratory, echinoplutei
of P. lividus can delay metamorphosis for 53 d after
reaching competence (Pedrotti 1990), but it is unknown how long larvae maintain competence in the
field. Even if metamorphosis takes place before settling (Fenaux & Pedrotti 1988), the post-larvae now
find themselves in a hostile environment and thus this
part of the population may be considered lost for benthic recruitment. The ophiuroid larvae studied have
better chances of recruitment. They have a planktonic
Life span longer than that of the echinoids (Pedrotti
1990).Metamorphosis precedes settling onto a surface
and these species are found at greater depths. However, according to Gage & Tyler (1981), post-larval
stages of the upper-bathyal Ophiocten gracilis, found
at 2900 m depth, are the consequence of dispersal to
non-favorable regions, and this represents a loss for
recruitment.
Coastal upwelling, according to its variation and intensity, may have a favorable or unfavorable effect on
larval transport in the surface waters. Along the coasts
of California and Oregon (USA), intense upwelling
advects competent larvae of the purple sea urchn
Strongylocentruspurpuratus away from the coast while
ed&es that form due to the cold water jets may bring larvae back to the coast (Ebert & Russell1988). In the California and Peru currents, upwelling and its associated
offshore transport acting previous to spawning activity
might enhance larval survival by ensuring adequate

226

Mar. Ecol. Prog. Ser.

food particle concentrations, and acting at the same

spatio-temporal scales as spawning could result in an
offshore loss of eggs and larvae (Bakun & Parrish 1982).
In the Ligurian Sea, the presence of a density front in
surface waters limits larval dispersal, while its absence
enhances the offshore transport of larvae.
The present study emphasizes the importance of
knowledge of physical processes in coastal regions to
understanding the ecology of larval echinoderms. The
influence of a hydrological barrier in distribution and
dispersal of echinoderm larvae raises questions about
the dynamics of other coastal meroplankton populations
in the Ligurian Sea. One may hypothesize that the relationship between upwelling intensity and larval production may be an important factor that determines
success in recruitment. However, the way by which
competent echinoderm larvae leave surface waters
to settle on the bottom still remains unknown. A close
survey of larvae near the bottom and a good description of circulation and mixing processes through the
water column are needed to elucidate this question.
Acknowledgements. We thank L. Prieur. S. Dallot. P. Nival,
M. Youngbluth, R. Scheltema and anonymous reviewers for
constructive comments, and A . M. Corre for help with plankton sample analysis. This work was supported by CNRS-UA
716 and FRONTAL (JGOFS/France) and PNDR programmes.

LITERATURE CITED
Bakun, A., Parrish, R. H. (1982).Turbulence, transport, and
pelagic fish in the California and Peru current systems.
CalCOFI Rep. XXXIII: 99-1 12
Banse, K. (1964). On the vertical distribution of zooplankton
in the sea. Prog. Oceanogr. 2: 53-125
Banse, K. (1986).Vertical distribution and horizontal transport
of planktonic larvae of echinoderms and benthic polychaetes in an open coastal sea. Bull. mar. Sci.
39(2): 162-175
Bethoux, J. P,, Nyffeler, F., Prieur, L. (1981). Utilisation des
m.oyennes hydrologiques pour le calcul des flux d'eaux
dans le bassin Liguro-provengal. Rapp. comm. i.nt, mer
Med. 27(6):193-196
Bethoux, J . P,, Prieur, L. (1983). Hydrologie et circulation en
Mediterranee Nord-Occ~dentale.Petrole et techniques
229: 24-34
Bethoux, J . P,, Prieur, L., Bong, J. H. (1988).Le courant Ligure
au large de Nice Oceanol. Acta 9: 59-67
Bong, J . H (1983) Contribution a l'etude du courant Ligure et
de sa variabilite spatio-temporelle a moyennes echelles.
These 3eme cycle, oceanographie physique, Univ. Paris
Boucher, J. (1984).Localization of zooplankton population in
the Llgurian marine front- role of ontogenlc migration.
Deep Sea Res. 31(5):469-484
Boucher, J . , Ibanez, F., Prieur, L. (1987). Daily and seasonal
variations in the spatial dlstributlon of zooplankton population m relatlon to the physical structure in the Ligurian
Sea Front. J. mar. Res. 45: 133-173
Bougis, P., Carre, C. (1960). Conditions hydrologiques a
Villefranche-sur-Mer pendant les annees 1957-1958.
Cahiers ocean. 12(6):393-408

Nival, P. (1988).La croissance et

Braconnot, J. C., Choe, S. M,,
le developpement de Salpa fusiforrnis Cuvicr (Tunicata,
Thaliacea). Arm. Inst. oceanogr. Paris 64: 101-114
Chia, F. S. (1974).Classificat~onand adaptative significance
of developmental patterns in marine invertebrates.
Thalassia jugosl. 10(1/2): 121-130
Choe, S. M. (1985). Contribution a I'etude biologique et
ecologique des Tunicens pelagiques Salpides (Thaliaces)
e n Mediterranee. These de doctorat 3Pme cycle. Univ.
Paris V1
Dallot, S . , Pizay, M. D. (1989). Changement des conditions
hydrologiques en mer Ligure entre mars 1988 et mars
1989. In: L'Helgven, S. (ed.) Devenir de la matiere organique produite dans les zones frontales, consommation,
mineralisation exportation, importance des phenomenes
physiques. 'Table Ronde' Frontal, INSV, Brest. 1 p.
Ebert, T. A.. Russel. M. P. (1988). Latudinal variation in size
structure of the west coast purple urchin: a correlation
with headlands. Limnol. Oceanogr. 33(2):286-294
Emlet, R. B. (1986). Larval production, dispersal, and growth
in a fjord: a case study on larvae of sand dollar Dendraster
excentricus. Mar. Ecol. Prog. Ser. 31- 245-254
Fenaux, L., Claustre, H., Cellario, C. (1987). Developpement
larvaire de Paracentrotus lividus (Larmak) dans des conditions de bloom phytoplanctonique. In: 5' Seminaire
International des Echinodermes actuels et fossils. Bull.
Soc. Sci. nat. Ouest France, Suppl. hors-serie, p. 83
(abstract)
Fenaux, L., Pedrotti, M. L. (1988).Metamorphose des larves
d'Ech~nidesen pleine eau. P.S.Z.N. I: Mar. Ecol. 9(2):
93-107
Gage, J. D., Tyler, P. A. (1981). Non-viable seasonal settlement of larvae of the upper bathyal brittle star Ophiocten
g r a c d ~ sin the Rockall Trough abyssal. Mar. Biol. 64:
153-161
Gascard, J. C. (1991). Open ocean convection and deep
water formation revisited in the Mediterranean, Labrador, Greenland and Weddel seas. In: Chu, P. C.. Gascard, J. C. (eds.) Deep convection and deep water formation in the oceans. Elsevier, Oceanographic Series 57, p.
157-181
Gostan, J. (1968). Contribution a l'etude hydrologique du
bassin Liguro-provengal entre la Riviera et la Corse. These
doctorat Etat, fac. sci. Univ. Paris
Greenblatt, P. R. (1982).Small-scale horizontal distributions of
zooplankton taxa. Mar. Biol. 67: 97-11
Haury, L. R. (1976).Small-scale pattern of a California current
zooplankton assemblage Mar. Blol 37. 137-157
Haury, L. R., McGowan, J. A., Wiebe, P. H. (1978). Patterns
and processes in the time-space scales of planktonic distributions. In: Steele, J. H. (ed.) Spatial pattern in plankton
communities. Plenum Press, New York, p. 277-327
Haury, L. R., Wiebe, P. H. (1982). Fine-scale multi-specles
aggregation of oceanic zooplankton. Deep Sea Res.
29(7A).915-921
Hentsch. J. M. (1959). Etude des courants de la Baie de
Villefranche. Master's thesis, Universite de Paris
Korringa, P. (1952) Recent advances in oyster biol.ogy. Q. Rev.
Biol. 27: 266-308 & 339-365
Mackas, D. L., Denman, K. L., Abbott, M. K. (1985).Plankton
patchiness: biology in the physical vernacular. Bull. mar
Sci. 37(2):652-674
Mileykovsky, S. A. (1966). Range of d~spersion of the
pelagic larvae of benthic invertebrates by currents and
the migratory role of this dispersion taking Gastropoda
and Lamellibranchia as examples. Oceanology 6(3):
396-404

Pedrotti & Fenaux: Dispersal of echinoderm larvae

227

Mileikovsky, S. A. (1968).Some common features in the drift

of pelagic larvae and juvenile stages of bottom invertebrates with marine currents in temperate regions. Sarsia
34: 209-216
Mileikovsky, S. A (1970) Seasonal and daily dynamics in
pelagic larvae of marine shelf bottom ~nvertebratesin
nearshore waters of Kandalaksa Bay [White Sea). Mar.
Biol. 5: 180-194
Mileikovsky, S. A. (1971). Types of larvae development in
marine botton invertebrates, their distribution and ecological significance: a re-evaluation. Mar. Biol. lO(3):193-213
Mileikovsky, S. A . (1974). Types of larval development in
marine botton invertebrates: an integrate ecological
scheme. Thalassia jugosl 10(1/2):171-179
Mortensen, Th. (1927).Handbook of the echinoderms of the
Bntish Isles. Oxford Un~v.Press, Oxford
Muller, J. (1852). Uber die Ophlurenlarve des Adnatischen
Meeres. Dumrnler, Berlin
Nival, P., Malara, G., Chara, R. (1975).Evolution du plancton
dans la baie de Villefranche-sur-Mer a la fin du printemps
(mai et juin 1971). I. Hydrologie, sels nutritifs, chlorophylle. Vie Milieu 25(2):231-260
Okubo, A. (1980). Diffusion and ecological problems: mathematical models. Biomathematics, Vol. 10. Springer-Verlag,
Berlin, p. 25-42
Pedrotti, M. L. (1990). Etude des processus biologiques et facteurs physiques responsables de la dispersion et du recrutement des larves meroplanctoniques. Modele: les
larves d'Echinodermes. These de doctorat de 1'Universite
Pierre et Marie Curie en Sciences de la Vie, Paris
Pennington, J T., Emlet, R. B. (1986). Ontogenetic and die1
migration of a planktonic echinoid larva Dendraster excentricus (Eschscholtz).occurrence, causes and probable
consequences. J. exp. mar. Biol. Ecol. 104: 69-95
Prieur, L. (1981). HBterogeneite spatio-temporelle dans le
bassin Liguro-provengal. Rapp. C o n ~ m .int. mer Med.
27(6): 177-179
Prieur, L., Raunet, J., Taillez, D. (1990). Radiale PROS V1
1981-1989. Rapport interne L.P.C.M. (Laboratoire de
Physique et Chimie Marines). CNRS, Villefranche-surMer
Shanks, A. L. (1983). Surface slicks associated with tidally
forced internal waves may transport pelagic larvae of benthic invertebrates and fishes shoreward. Mar. Ecol. Prog.
Ser 13: 311-315
Shanks, A. L. (1985). Behavioral basis of internal-waveinduced shoreward transport of megalopae of the crab
Pachygrapsus crassipes. Mar. Ecol. Prog. Ser. 24: 289-295
Shanks, A. L. (1986). Tidal periodicity in the daily settlement
of intertidal barnacle larvae and a hypothesized mechanism for the cross-shelf transport of cyprids. Biol. Bull. 170:
429-440
Shanks, A. L., Wright, W. G . (1987). Internal wave mediated
shoreward transport of cyprids megalopae, and gammarids and correlated longshore differences in the settling
rate of intertidal barnacles. J exp. mar. Biol. Ecol. 114:
1-13
Scheltema, R. S. (1966). Evidence for trans-Atlantic transport
of gastropod larvae belonging to the genus Cyrnatium.
Deep Sea Res. 13: 83-95

Scheltema. R. S. (1968). Dispersal of larvae by equatorial

ocean currents and its importance to the zoogeography of
shoal-water tropical species. Nature, Lond. 217:
1159-1162
Scheltema, R . S . (1971).The dispersal of the larvae of shoalwater benthic invertebrate species over long distances by
ocean currents. In: Crisp, D J. (ed.) Proc. 4th Eur. Mar.
Biol. Symp. Cambridge Univ. Press. London, p. 7-28
Scheltema, R. S. (1972). Dispersal of larvae as a means of
genetic exchange between widely separated populations
of shoal-water benthic invertebrate species. In: Battaglia,
B. (ed ) Proc. 5th Eur. Mar. Biol. Symp. Piccin, Padua, p.
101-144
Scheltema, R. S. (1974). Biological interactions determining
larval settlement of marine invertebrates. Thalassia jugosl.
10(1/2):263-296
Scheltema. R. S. (1975).The frequency of long-distance larval
dispersal and the rate of gene-flow between widely separated population of sipunculans. In: Rice, M. E., Todorovic,
M. (eds.) Proc. of International Symposium on the Biology
of the Sipuncula and Echiuna. Vol. 1. Institute for
Biological Research, Belgrade, and National Museum of
Natural History, Smithsonian Institution, Washington,
D.C., p. 18-25
Scheltema, R . S. (1986).On dispersal and planktonic larvae of
benthic invertebrates: an eclectic overview and summary
of problems. Bull. mar. Sci. 39(2): 290-322
Scheltema. R. S. (1988). Initial evidence for the transport of
teleplanic larvae of benthic invertebrates across the east
Pacific Barrier. Biol. Bull. 174: 145-152
Scheltema, R. S . , Rice, M. E. (1990). Occurrence of teleplanic
pelagosphera larvae of sipunculans in tropical regions of
the Pacific and Indian oceans. Bull. mar. Sci. 47(1):
159-181
Sournia, A., Brylinsky, J. -M,, Dallot, S., Le Corre, P,, Levau,
M., Prieur, L., Froget, C. (1990) Fronts hydrologiques au
large des cotes fran~aises:les sites-atelies du programme
Frontal. Oceanol. Acta 13: 413-438
Strathmann, R. R. (1975).Larval feeding in echinoderms. Am.
ZOO^. 15: 717-730
Thorson, G . (1961).Length of pelagic larval life in marine bottom invertebrates as related to larval transport by ocean
currents. In: Sears, M. (ed.) Oceanography. Am. Ass.
Advmt. Sci., Washington, D.C., 67. 455-474
Tranter, D J. (1967).A formula for the filtration coefficient of
a plankton net. Aust. J. mar. Freshwat. Res. 1. 113-121
Vance, R. R. (1973a). On reproductive strategies in marine
benthic invertebrates. Am. Nat. 107: 339-351
Vance, R. R. (1973b). More on reproductive strategies in
marine benthic invertebrates. Am. Nat. 107: 353-361
Wieghardt, K . E. G. (1953). On the resistance of screens.
Aeronaut. Q. 4: 186-172
Yamamoto, T., Nishizawa, S. (1986).Small-scale zooplankton
aggregations at the front of a Kuroshio warm-core ring.
Deep Sea Res. 33(11/12):1729-1740
Young, C. M., Chia, F. (1987).Abundance and distribution of
pelagic larvae as influenced by predation, behavior, and
hydrographic factors. In: Giese, A., Pearse, J., Pearse, V. B.
(eds.) Reproduction of marine invertebrates, Vol. 9.
Blackwell Sci. Publ., Palo Alto, California, p. 385-463

This article was presented by R. S. Scheltema, Woods Hole,

Massachusetts, USA

Manuscript first received: December 5, 1991

Revised version accepted: June 25, 1992