Eur. J. Phycol. (2005), 40(2): 149–158
Effects of wave exposure and depth on biomass,density and
fertility of the fucoid seaweed Sargassum polyceratium
ASCHWIN H. ENGELEN
,JEANINE L. OLSEN
WYTZE T. STAM
AND ANNEKE M. BREEMAN
Centre for Ecological and Evolutionary Studies (CEES),Department of Marine Biology,University of Groningen,
P.O. Box 14,9750 AA Haren,The Netherlands
Department of Marine Ecology,Marine Botany,University of Go
¨teborg,Box 461,405 30 Go
(Received 3 September 2004; accepted 18 February 2005)
Sargassum polyceratium is widely distributed around the island of Curac¸ ao (Netherlands Antilles) where it inhabits strongly
contrasting habitats. Changes in population structure have been followed in three bays with increasing levels of wave exposure
at two depths: shallow (0 m) and deep (18 m). The effects of increasing wave exposure were investigated by studying three deep-
water populations; and the effects of depth by studying shallow- and deep-water populations in the two calmer bays. Total
density and stage density (reproductive thalli, juvenile thalli) were determined and total and individual thallus biomass was
estimated non-destructively. In the most wave-exposed deep-water population thalli were twice as long with more than twice
the biomass than in the calmest deep-water population. Total density and juvenile density were highest in the bay with
intermediate wave exposure. Depth was an important factor at both the individual and population level. Shallow-water thalli
had basal holdfast areas that were four times larger than those from deep water, and thallus biomass was positively correlated
with holdfast area. Shallow-water juveniles invested more in the development of a holdfast and lateral growth than deep-water
juveniles. Total biomass per quadrat was up to 10-fold higher in shallow- than in deep-water populations. In shallow-water
populations reproductive thalli were present throughout the year whereas in deep-water populations they were present only
during autumn and winter. We conclude that both wave exposure and depth affect population structure. Thalli were generally
bigger and total biomass higher in the more exposed bay(s) and in shallower water, contradicting the general pattern in
macroalgae of reduced size and biomass with increasing wave exposure.
Key words: biomass, density, depth, fertility, population structure, Sargassum polyceratium, thallus size, wave exposure
Spatial variability is a fundamental characteristic
of animal and plant populations, in part reflecting
variability in the environment itself. Spatial and
temporal heterogeneity have major consequences
for processes such as reproduction, competition
and predation (Levin, 1992) and thus, need to
be considered in some detail (Kareiva, 1990).
As a starting point for such studies, habitat-related
differences in biomass, density, stage structure
and fertility need to be documented, prior to
investigating questions related to population
Sargassum polyceratium is a fucoid seaweed
that inhabits a broad spectrum of wave exposure
environments throughout the Caribbean. It occurs
from southern Florida to northern Brazil
(Diaz-Piferrer, 1969) and is widely distributed in
shallow as well as deep water in the Lesser
Netherlands Antilles (Van den Hoek, 1969; Van den
Hoek et al., 1975; Vermeij, 1997; Engelen, 2004).
Around the island of Curac¸ ao, it is a dominant
macrophyte in the wave-swept intertidal zone on
both the heavily exposed windward coast and
protected leeward coast. It also occurs in deeper
reef zones on both leeward and windward shores
(Engelen, pers. obs.).
Seasonal patterns of variation with respect to
depth were studied by De Ruyter van Steveninck
and Breeman (1987). They quantified changes in
biomass, growth rate, turnover of primary laterals
and fertility of an intertidal (0 m) and a deep-water
Correspondence to: Aschwin H. Engelen. Present address:
CCMAR, Universidade do Algarve, FCMA, Campus de
Gambelas, 8005-139 Faro, Portugal. Fax: 00531-289818353.
ISSN 0967-0262 print/ISSN 1469-4433 online ß2005 British Phycological Society
(30 m depth) population of S. polyceratium in one
leeward bay on Curac¸ ao. Intertidal thalli had a
more compact growth form, faster growth rate,
more rapid turnover of laterals and remained
fertile throughout the year. The subtidal thalli
were more elongated, grew more slowly and were
fertile only seasonally. These authors concluded
that depth (and by extension probably light
limitation) was the main habitat variable account-
ing for their observations. Later observations from
other bays with different exposures (by Engelen
and colleagues around the island) suggested that
wave action, and not just depth, may be a more
important factor affecting standing biomass,
thallus size, morphology and, possibly, fertility.
Wave action is known to be one of the primary
factors affecting population structure and
dynamics in benthic marine habitats (Gaylord,
1999). In addition, morphological changes in
seaweeds experiencing heavy waves and surge are
well documented (see reviews by Norton et al.,
1981; Norton, 1991). Algae growing in exposed
locations are typically smaller, thicker, more
strongly branched and/or have stronger holdfasts
than those growing in calmer waters (Norton et al.,
1981; Norton, 1991; Sjøtun & Fredriksen, 1995;
Hurd, 2000). The general trend of decreasing
thallus size with increasing wave exposure has
also been observed in Sargassum (De Paula &
Oliveira Filho, 1982). Although the advantage of
such wave-exposed morphologies in reducing
hydrodynamic drag and preventing dislodgment
is well established (see Norton, 1991; Hurd, 2000
for reviews), the physiological and ecological
consequences of different wave exposure regimes
have been a matter of some debate. The view that
nutrient diffusion across the diffusion boundary
layer would be limited in calm environments
(Wheeler, 1980, 1988) has recently been questioned
by Hurd (2000). Blanchette (1997) found that,
under exposed conditions, the mean size of Fucus
gardneri was reduced by reduction of thalli of
all sizes rather than by the loss of larger thalli.
The aim of the present study was to quantify
differences in biomass, density and stage structure
of S. polyceratium in contrasting environments.
We first investigated the effect of wave exposure by
comparing three populations along an increasing
wave exposure gradient around the island at a
constant depth of 18 m. Next, we investigated the
effect of depth by comparing shallow-water popu-
lations (0 m) with deep-water populations (18 m) in
the two calmer bays.
The following questions were addressed at the
individual thallus level: (i) Do the sizes of adults
and juveniles vary with wave exposure and/or
depth? (ii) Are there depth-related differences in
The following questions were addressed at the
population level: (i) Do populations differ in total
density, juvenile density and biomass per unit area
with wave exposure and/or depth? (ii) Are tempor-
al changes in density, biomass and reproduction
dependent on depth?
Material and methods
The research was conducted between June 1996 and
June 1997 on the island of Curac¸ ao (Netherlands
Antilles), 60 km off the coast of Venezuela (Fig. 1).
Background information about the area can be found in
Van den Hoek (1969), Van den Hoek et al. (1975) and
Van Duyl (1985). Information on intertidal and subtidal
vegetation can be found in Van den Hoek (1969),
Wanders (1976), and Van Loenhoud & Van de Sande
(1977). Detailed descriptions of the coral reef compo-
nents can be found in Bak (1977), Van den Hoek et al.
(1975, 1978) and Van Duyl (1985).
Populations of Sargassum polyceratium Montagne
were studied in three bays with different degrees of wave
exposure (Fig. 1). St. Michiel Bay, on the leeward
southwest coast, receives the least exposure, with waves
from 0–30 cm high. Westpunt Bay, on the northwest tip
of the island has only slightly higher waves (0–50 cm) but
is also subjected to a refracted oceanic swell that wraps
around the tip. This can bring wave heights to 1 m.
Canoa Bay, on the windward northeast coast, is the most
wave-exposed bay with waves from 2–3.5 m high (Van
The long-term average (1964–1980) wind velocity on
Curac¸ ao is 7.1 m s
, with highest values occurring in
January–August (7–8 m s
), and somewhat lower
Fig. 1. Map of Curac¸ ao, Netherlands Antilles, with
location of bays in which populations of Sargassum
polyceratium were studied: St. Michiel Bay (M), Westpunt
Bay (W), and Canoa Bay (C) and diagram of sampling
design used to study effects of wave exposure and depth.
A. H. Engelen et al. 150
values in September–December (6–7 m s
). In the years
of study, the average wind velocities were somewhat
lower than the long-term average (6.0 and 6.26 m s
1996 and 1997, respectively), and mean monthly values
ranged from 4.6 to 7.4 m s
(data courtesy of the
Meteorological Service of the Netherlands Antilles and
The mean daily tidal range at Curac¸ ao is small
(29.7 cm 10.7 [SD]; De Haan & Zaneveld, 1959).
Intertidal populations of S. polyceratium along the
leeward coast are almost constantly exposed to gentle
waves. Total water motion measured with gypsum
blocks indicated a 10-fold higher water motion in the
intertidal zone than in the subtidal zone in leeward bays
(Engelen, unpublished data). Intertidal populations of S.
polyceratium along the windward coast are constantly
exposed to heavy waves and are mostly inaccessible.
Michiel Bay is about 300 m wide. Intertidal
S. polyceratium is present only in the western part of
the bay. It occurs in a belt 100 m long and 1–4m wide on
a limestone terrace that slopes gradually from sea level
to a depth of 0.5 m. Only the intertidal part of the bed
was studied. S. polyceratium is absent from the reef flat
where the substratum is covered by sand. At 10 m depth,
a rather steep drop-off begins. S. polyceratium is found
on the reef slope from 12 m down to at least 30 m.
Westpunt Bay is about 500 m wide. An intertidal
terrace composed of natural limestone and artificial
concrete is restricted to the northern part of the bay.
Here, the belt of S. polyceratium is 100 m long and 2–3 m
wide. Depth in front of the terrace is approximately 0.5–
1 m. In the subtidal zone, between 1 and 10 m depth, the
substratum is sand and coral rubble (dead Acropora
cervicornis). At 10–12 m depth, a gentle slope begins and,
at 25–30 m depth, coral formations are replaced by sand.
Compared with other leeward sites on Curac¸ ao, the reef
at Westpunt is well developed, being dominated by
corals. S. polyceratium is found both on the reef flat and
the reef slope down to about 60 m depth.
Canoa Bay is about 300 m wide with a narrow
S. polyceratium belt on the intertidal limestone terrace
and a broad meadow in the subtidal zone. Continuous
heavy wave action on the windward coast prevented
intertidal work, and access to the extensive subtidal
Sargassum beds was possible only during calm weather.
This occurs on some days during the Caribbean
hurricane season from autumn to early winter, when
the easterly trade wind switches to the west. During the
study period, sampling of subtidal Sargassum beds was
possible on only two occasions in November and
Sampling design and measurement of variables
A summary of the sampling design is presented in Fig. 1.
The effect of wave exposure was studied over a wave
exposure gradient (all three bays) at 18 m depth in
November and December (1996). The effect of depth
was studied by comparing intertidal (0 m) and deep,
subtidal (18 m) populations in the two calmer bays.
These were typically 100–150 m apart. Shallow- and
deep-water populations in the two bays were sampled at
monthly intervals between October and April. Within
each depth, two sites of 30 m
each and 25–50 m apart
were marked. Within each site, at least five quadrats of
were placed at random positions during each
census. Where densities were low, the number of
quadrats was increased to provide sufficient replication
of individual thallus parameters (see below).
In every quadrat, the length and maximum circum-
ference of each thallus were measured; and the total
number of thalli, number of reproductive thalli and
number of juveniles were counted. Individuals were
considered reproductive when receptacles were present.
Juveniles were defined as small rosettes of leaf-like
fronds without a detectable holdfast, main axes or
laterals. Juveniles could be detected when about 0.5 cm
A non-destructive estimate of the biomass of each
thallus, based on length (L) and maximum circumfer-
ence (C), was made at each census using the method of
˚berg (1990). This was done by regressing the parameter
against thallus dry weight. Both length and
maximum circumference were measured with a flexible
measuring tape to an accuracy of 0.5 cm. Thalli with a
circumference smaller than 2.0 cm could not be mea-
sured accurately and their circumference was defined as
2.0 cm. To obtain the relationship between LC
weight, other thalli were harvested from each location in
different months, covering the whole study period.
Length and maximum circumference were measured
on 552 individuals, together with their wet and dry
weights. Dry weights were determined after drying for
two days at 80C. Wet weight and dry weight were
determined to a precision of 0.01 g. The dry weight:wet
weight ratio was 0.19. Because of their skewed distribu-
and dry weight were natural log-transformed.
An analysis of covariance (ANCOVA) was used to
detect differences between sampled populations and
dates in order to pool some of the data. A significant
difference in the slope of the regression was found
between shallow and deep-water samples ( p¼0.006).
The predictive equation for the shallow-water samples
ðr¼0:96, N¼207, p<0:001Þ;
and for the deep-water samples:
ðr¼0:92, N¼345, p<0:001Þ;
in which DW is thallus dry weight, L is thallus length
and C is the maximum circumference.
To test for differences in holdfast area between
shallow- and deep-water thalli and to test for a possible
correlation between thallus size and holdfast area, 90
thalli were randomly sampled from shallow- and deep-
water populations in the two calmer bays. The dry weight
of each thallus was measured and the basal holdfast area
Effects of wave exposure and depth on Sargassum polyceratium 151
estimated by multiplying the maximum width by the
width perpendicular to the maximum. The relationship
between basal holdfast area and thallus dry weight was
analyzed with a Model 1 regression.
Since substratum area can be increased by relief,
differences in substrate relief between deep and shallow
sites were examined by estimating the surface area under
ten 0.25 m
quadrats at each site. The distances between
adjacent corners of each quadrat were measured over
the underlying substratum with a flexible measuring
tape. This resulted in two X and two Y substrate surface
distances per quadrat. Quadrat substrate area was
calculated as 0.50 (X1Y1 þX2Y2). Analysis of variance
showed that substrate area was smaller at Michiel than
at Westpunt ( p¼0.006), although the difference was
small (8%). Post hoc pooling of the non-significant
interaction term between bay and depth showed that
substrate area was about 80% greater in deep water
) than in the intertidal zone (0.315 m
For each quadrat, the following variables were analyzed:
thallus density, juvenile density, proportion of repro-
ductive thalli (relative to the total number of adult thalli
excluding juveniles and bare holdfasts), thallus length,
juvenile length, estimated thallus dry weight, and total
biomass per quadrat.
Individual thallus variables were analyzed per quadrat
and quadrat variables were analyzed per site. As the
number of sampled quadrats was increased where
densities were low, random subsamples were taken
from the total data set in order to maintain a balanced
sampling design in the analysis of variance (ANOVA).
For the comparison of deep sites with different degrees
of wave exposure, mean thallus variables were calculated
from 5 randomly chosen thalli per quadrat and 5
randomly chosen quadrats per site. Quadrats with <5
thalli were excluded before random subsampling. Mean
juvenile length was calculated from 2 randomly chosen
juveniles per quadrat and 2 randomly chosen quadrats
per site. Mean population variables were calculated
from 5 randomly chosen quadrats per site, involving
random subsampling among all quadrats. Due to
occasionally low densities at one of the sites, the
number of quadrats available for analyses of individual
thallus variables in the comparison of shallow- and
deep-water populations in the two calmer bays had to be
lowered to 3 quadrats per site (excluding quadrats with
<3 thalli before random subsampling) with 3 randomly
chosen thalli per quadrat. Mean juvenile length was
calculated as before. Mean population variables were
calculated from 6 randomly chosen quadrats per site,
involving random subsampling among all quadrats.
Variation in thallus and population variables was
analyzed with ANOVA in which Bay, Depth and
Month were considered as (orthogonal) fixed factors
and Sites and Quadrats were considered as (nested)
random factors. Significant interaction terms of fixed
factors were further analyzed according to Underwood
(1997). The assumption of homogeneity of variances
was analyzed with Cochran’s C-test (Winer et al., 1991).
Where variances were heterogeneous, data were
transformed. In general, data on individual length and
dry weight were natural log-transformed. Data on
thallus numbers were first square root transformed; if
not successful in reducing heterogeneity, values were
raised to a smaller power. Transformations were not
always successful in homogenizing variances. As sug-
gested by Underwood (1997), the analysis was then
performed on untransformed data, since the datasets
were relatively large and balanced. Multiple means
comparisons were made with the Student-Newman-
Keuls test (SNK) and variance components were
calculated as described by Underwood (1997).
Deep-water populations over a wave exposure
Mean overall thallus length varied significantly
among bays (ANOVA, p¼0.009; Fig. 2A).
At Canoa Bay, thalli were about twice as long as
at Westpunt and Michiel. Juvenile lengths differed
only at the small spatial scale of sites (ANOVA,
p¼0.045), mean lengths ranged from 1.38–4.19 cm
(data not shown).
Thallus biomass (estimated dry weight) was
lower in the calmest bay (Michiel) than in the
two more exposed bays (ANOVA, p¼0.040;
Fig. 2B). Small-scale spatial variation was found
among quadrats (ANOVA, p¼0.001).
No significant difference in the proportion
of reproductive thalli was found among bays
(data not shown) but there was small scale spatial
and temporal variation (ANOVA, bay: p¼0.230,
month*site: p¼0.002). The proportion of
reproductive thalli was, on average, 11.4%.
Total densities were highest at Westpunt, inter-
mediate in the windward bay of Canoa and lowest
in the most sheltered bay of Michiel (total density
ANOVA, p¼0.008; Fig. 2C); juvenile densities
were higher at Westpunt than at Michiel (juvenile
density ANOVA, p¼0.044; Fig. 2C). Total density
also showed small scale spatial and temporal
variation (ANOVA, month*site: p¼0.002).
Total dry weight per quadrat was an order of
magnitude higher at Canoa and Westpunt than
at Michiel (ANOVA, p¼0.004; Fig. 2D) and also
showed small scale spatial and temporal variation
(ANOVA, month*site: p¼0.018).
Comparing shallow- and deep-water populations
in the two calmer bays
Thallus size. No significant differences in thallus
length were detected between depths but there was
a significant month*bay interaction (ANOVA,
p¼0.008; Fig. 3). From November to March
thalli were longer (1.7–3.2 times) at Westpunt
than at Michiel. Small-scale spatial variation was
A. H. Engelen et al. 152
found among sites (ANOVA, p¼0.009) and
quadrats (ANOVA, p¼0.015).
Juveniles were more than twice as long at 18 m
(2.3 cm) as in the intertidal zone (1.0 cm) (ANOVA,
p¼0.014). Juveniles were about 25% longer at
Westpunt (1.9 cm) than at Michiel (1.5 cm)
(ANOVA, p¼0.019). Temporal variation was
detected, but depended on site (ANOVA,
Biomass was significantly lower in deep than in
shallow water, at least during some months
(ANOVA, month*depth: p¼0.037; Fig. 4).
Significant temporal variation was present in
shallow but not in deep-water thalli. There was
also significant variation among sites (ANOVA,
month*site: p¼0.042) and quadrats (ANOVA,
Holdfasts were almost 4 times larger in the
intertidal zone (1.8 cm
) than at 18 m depth
; ANOVA, p¼0.0001). Holdfast area was
significantly correlated with thallus dry weight, but
the relationship was stronger in the intertidal zone
¼0.392, p<0.001) than in deep water
Reproduction. The proportion of reproductive
thalli depended on depth and bay with varying
temporal patterns (ANOVA, month*bay*depth:
p¼0.001; Fig. 5). There was also significant
Mean thallus dry weight (g)
Mean total dry weight (g 0.25 m-2)
Mean density (0.25 m-2)
Mean thallus length (cm)
Fig. 2. Thallus length (A), estimated thallus dry weight
(B), total (hatched bars) and juvenile (black bars) densities
(C), and total dry weight per quadrat (D) of Sargassum
polyceratium at 18 m depth in three bays (Michiel,
Westpunt and Canoa), arranged in order of increasing wave
exposure. Bars indicate standard error (A, B; n¼100) or
standard deviation (C, D; n¼20); lower case letters indicate
grouping of means (means which have one or more letters
in common are not significantly different at p¼0.05).
Oct Nov Dec Jan Feb Mar Apr
Mean thallus length (cm)
Fig. 3. Thallus length of Sargassum polyceratium over time
in the two calmer bays (Westpunt and Michiel; data
combined for 0 and 18 m depth according to ANOVA
results). Bars indicate standard error (n¼36); black bar
above indicates significant difference between bays; other
details as in Fig. 2.
Fig. 4. Estimated thallus biomass of Sargassum polycer-
atium over time in shallow- (0 m) and deep-water (18 m)
populations in the two calmer bays (Westpunt and Michiel;
data combined for the two bays according to ANOVA
results). Bars indicate standard error (n¼36); black bars
above indicate significant difference between depths; other
details as in Fig. 2.
Effects of wave exposure and depth on Sargassum polyceratium 153
variation among sites (ANOVA, p¼0.001). At 0 m
depth, reproductive thalli were present throughout
the year, and their proportion remained constant
in both bays. At 18 m depth the first reproductive
thalli were detected in October at Michiel, but they
were found in the sampled quadrats only from
November/December to April (Fig. 5). At Michiel,
the proportion of reproductive thalli was consis-
tently low and showed no significant temporal
variation. At Westpunt, reproductive thalli first
appeared in December; their proportion had
increased by January and remained stable to
April (Fig. 5). In May, a few reproductive thalli
were still present but they were absent during the
summer months. During months when reproduc-
tion occurred at both depths, the proportion of
reproductive thalli was generally similar at 0 and
18 m (Fig. 5). At both depths, the proportion of
reproductive thalli was higher at Westpunt than at
Michiel (0 m: December and January; 18 m:
January–April; ANOVAs: p<0.05; Fig. 5).
Density and total biomass per quadrat. No
significant difference in total density was detected
between depths but densities were 2.5-fold higher
at Westpunt (102.8 m
) than at Michiel (43.2 m
(ANOVA, p¼0.0001). In addition there was small
scale spatial and temporal variation (ANOVA,
month*site: p¼0.027). Juvenile densities varied
only among sites (ANOVA, p¼0.016). On average
16.7 juveniles (m
) were counted.
The total biomass per quadrat differed with
depth and bay, with varying temporal patterns
(ANOVA, month*bay*depth: p¼0.015; Fig. 6).
Total biomass was higher at 0 m than at 18m depth
in all months and in both bays (Single factor
ANOVAs, p-values ranging from <0.001 to
0.043). Total biomass was also consistently higher
at Westpunt than at Michiel (Single factor
ANOVAs, p-values ranging from <0.001 to
0.029). Significant temporal variation was found
at 18 m depth at Michiel and at 0 m depth at
Westpunt (Fig. 6).
Does thallus size differ depending on the level of
wave exposure and/or depth?
In many algal species, thalli growing in the
intertidal zone, exposed to higher wave action,
are tougher and shorter than those in the subtidal
zone (see Norton et al., 1981; Norton, 1991; Hurd,
2000). This was, however, not true for Sargassum
polyceratium. Overall, thalli had the greatest
biomass under high wave exposure regimes.
Oct Nov Dec Jan Feb Mar Apr
Mean proportion of reproductive thalli
Fig. 5. Proportion of reproductive thalli of Sargassum
polyceratium over time in shallow-(0 m) and deep-water
(18 m) populations in the two calmer bays (Michiel and
Westpunt), relative to the total number of adult thalli per
quadrat. Bars indicate standard deviation (n¼12); black
bars within panels indicate significant difference between
depths; other details as in Fig. 2.
Fig. 6. Total dry weight per quadrat of Sargassum
polyceratium over time in shallow-(0 m) and deep-
water (18 m) populations in the two calmer bays
(Michiel and Westpunt). Note difference in scale of
y-axes; details as in Fig. 5.
A. H. Engelen et al. 154
High wave exposure has potentially opposite
effects on thallus biomass. Although it results in
better-developed attachment structures, which sup-
port bigger thalli (see below), it also causes lateral
loss or breakage, which leads to a reduction in
biomass (e.g., Blanchette, 1997). In S. polyceratium,
growth rates at the more exposed locations are
apparently high enough to counterbalance the
higher loss rates. Nevertheless, S. polyceratium
thalli at Curac¸ ao are considerably shorter (on
average 10–12 cm) than has been reported for
other locations (45–90 cm: Taylor, 1960; up to
36 cm: Prince, 1980).
As suggested by De Ruyter van Steveninck and
Breeman (1987) subtidal thalli may be light-limited
(at 18 m depth only one third of surface PAR
remains; Boelen, pers. comm.). However, reduced
water motion probably plays a more important
role than light limitation because algae at 18 m
depth were about twice as long in the most exposed
bay as in the calmer bays (Fig. 2A). Decreased
water motion is known to reduce gas exchange,
nutrient uptake and growth (see Wheeler, 1988;
Norton, 1991). Although Hurd (2000) recently
questioned whether mass transfer across the
diffusion boundary layer (DBL) would be limiting
in situ (constantly changing DBL thickness would
limit mass transfer only on a scale of seconds,
even under calm conditions), decreased water
motion might still cause nutrient limitation in the
oligotrophic waters around Curac¸ ao.
Juveniles of S. polyceratium were much longer in
deep than in shallow water. Our results suggest a
difference in growth strategy between shallow- and
deep-water juveniles, involving an earlier transition
into the adult stage in shallow water. In the well-lit
intertidal zone, with high wave exposure and high
thallus densities, it is probably of competitive
advantage to invest in holdfast development (to
protect against future biomechanical stress) and
growth of upright axes. In the more shaded and
less dense deep-water populations it might be a
better strategy to invest in light-capturing blades.
It is uncertain whether this is a common strategy
among Sargassum juveniles since most studies have
used size classes to distinguish juveniles from adult
individuals (e.g., Ang, 1991). The use of this size
class approach masks possible differences in
juvenile growth strategy. An experimental labora-
tory study by Hwang and Dring (2002) has shown
that the transition from juvenile to adult is under
photoperiodic control in S. muticum.
Are there depth-related differences in holdfast size?
The surface area of holdfasts was about 4-fold
larger in shallow- than in deep-water thalli. Higher
wave exposure demands better attachment to
withstand higher hydrodynamic forces (Norton,
1991; Gaylord, 1999; Hurd, 2000). There was a
stronger correlation between thallus and holdfast
size in shallow than in deep water, suggesting
greater constraints on the relationship between
holdfast area and biomass in shallow water. Larger
holdfasts bear more primary axes (Engelen, 2004);
these, in turn, bear a higher number of primary
laterals (De Ruyter van Steveninck & Breeman,
1987), resulting in more strongly branched thalli
that are less susceptible to hydrodynamic drag and
have a higher surface to volume ratio, enhancing
nutrient uptake and growth.
Do populations differ in density and/or biomass
depending on the level of wave exposure and/or
Total densities in the subtidal zone (18 m) were 4–6
times higher at Canoa and Westpunt, than at the
calmest bay (Michiel). Differences in Sargassum
density with wave exposure have been reported
earlier by De Paula and Oliveira Filho (1982), who
found 4-fold higher densities in exposed than in
sheltered intertidal populations of S. cymosum in
Brazil. Although densities per quadrat did not
differ significantly between shallow- and deep-
water populations, the 80% higher surface relief in
deep water implies that densities per substrate
surface area are higher in the intertidal zone.
We found that total biomass per quadrat at 18 m
depth was about 15-fold greater at Canoa and
Westpunt than in the calmest bay (Michiel). Total
biomass was also approximately 10-fold greater in
shallow- than in deep-water populations. Other
authors have reported the reverse, i.e., greater
biomass in the subtidal than in the intertidal
zone (e.g., S. carpophyllum,S. ilicifolium and
S. siliquosum; Hurtado & Ragaza, 1999; Trono &
Tolentino, 1993). However, their subtidal popula-
tions were located in shallower water (8–11 m) and
therefore probably less influenced by light and
nutrient limitation (more water movement).
In S. polyceratium, higher densities as well as
heavier thalli contributed to the higher total
biomass in the intertidal than in the subtidal
zone. Differences in thallus architecture, especially
the density of leaf-like fronds, are also important.
This was apparent from the steeper slope of the
regression line (LC
vs. individual dry weight) used
to estimate individual size in shallow- compared to
Are temporal changes depth-dependent?
The most striking depth-dependent temporal
variation was found in reproductive seasonality
(also reported by De Ruyter van Steveninck &
Breeman, 1987). Shallow-water populations are
Effects of wave exposure and depth on Sargassum polyceratium 155
fertile throughout the year. In the intertidal zone at
Westpunt, for example, up to 80% of individuals
were fertile. Similar high values have been reported
for S. henslowianum (Ang, 2000), S. muticum
(Deysher, 1984), S. johnstonii,S. herporhizum and
S. sinicola (McCourt, 1984). Deep water thalli,
living under marginal conditions, reproduce
only seasonally. Apparently reproduction cannot
be sustained all year round because of severe
Differences in reproductive phenology are often
minimal in the tropics, but become more promi-
nent with latitude (in correspondence with tem-
perature and light cues). Latitudinal differences in
reproductive seasonality have been observed over
species distribution ranges. For instance, in
S. muticum, reproduction starts earlier and lasts
longer towards the southern end of the distribution
(Norton & Deysher, 1989; Yoshida, 1983;
Espinoza 1990). Similarly, shallow water popula-
tions of S. polyceratium in Florida reproduced only
in winter and early spring (Prince, 1980), whereas
reproduction continued throughout the year at
Curac¸ ao. However, S. polyceratium is a pseudo-
perennial in Florida that is reduced to its holdfast
stage in midsummer (Prince, 1980).
No temporal variation in juvenile density was
detected, either in shallow or deep water. Since
dispersal of zygotes is limited to ca. 1 m from the
parent individual (Kendrick & Walker, 1991, 1995;
Stiger & Payri, 1999), we expected to find temporal
variation in juvenile density in the subtidal zone
because reproduction is restricted to winter and
early spring. Kendrick and Walker (1994) found a
strong increase in juvenile density 2–3 months after
the onset of reproduction. The absence of season-
ality suggests that the time needed for development
of microscopic germlings into detectable juveniles
is variable. This could be explained by the presence
of a germling bank (Hoffmann & Santelices, 1991;
Creed et al., 1996a,b) that is restocked seasonally
in the deep-water populations, but proceeds to the
juvenile stage class slowly, due to small-scale
variability in growth conditions (for instance light
and grazing pressure). On the other hand, recruit-
ment may not only depend on local reproductive
patterns, but also occur through drifting fertile
laterals from nearby intertidal populations
(Engelen et al., 2001).
Scale issues,phenotypic plasticity vs. ecotypic
Patterns of spatial and temporal variation may
provide pointers to the mechanisms causing this
variation. Factors influencing spatial heterogeneity
are numerous. Almost all variables measured in
this study varied at the small spatial scales of sites
or quadrats. Variation at these scales (metres to
tens of metres) is partially stochastic and partially
related to natural heterogeneity of the substratum
(Underwood & Chapman, 1998; Coleman, 2002).
At the scales of depth and bay, differences in
abiotic (or biotic) factors or episodic events are
most likely to play key structuring roles (A
Depth-related differences in morphology and
phenology (holdfast, individual and juvenile size;
reproduction) may be due to either plasticity or
ecotypic differentiation. Engelen et al. (2001)
analyzed spatial variation in genetic population
structure around the island of Curac¸ ao using
random amplified polymorphic DNA (RAPD)
markers. Genetic differentiation was found
among bays and between depths. Thus, there is
some level of selection or restricted exchange
among bays and between shallow and deep water
within bays. Around the entire island, however, no
differentiation between shallow- and deep-water
populations was detected. Populations at Westpunt
were most closely related to northwestern wind-
ward populations, and those at Michiel were
closely related to southeastern leeward and wind-
ward populations. The deep-water populations of
Westpunt and Canoa were more closely related to
each other than to the population at Michiel
(Engelen et al., 2001). Therefore, the question of
ecotypic differentiation or phenotypic plasticity
remains unresolved at present. In principle, reci-
procal transplant experiments could resolve the
issue but, this is not possible in practice because of
high loss rates.
Thallus size, thallus architecture and population
structure of Sargassum polyceratium were affected
by both wave exposure and depth. Contrary to
patterns found in other seaweeds, thalli were
generally bigger, populations denser and total
biomass was higher in the more exposed environ-
ments and shallow water than in calmer and deep
water. Shallow- and deep-water thalli differ in
several aspects of their life history, including
the transition from the juvenile to the adult stage
class, holdfast size and branching, and the timing of
reproduction. These differences may be due to either
phenotypic plasticity, or ecotypic differentiation.
We thank M. Dring, G. Pearson and two
anonymous reviewers for helpful comments.
Many thanks to Ria Siertsema, Mark Vermeij,
Wobine de Sitter and Yvon Geurts for their
A. H. Engelen et al. 156
dedicated help with the fieldwork. We thank the
director and staff of the Ecological Institute
CARMABI (Curac¸ ao) for working facilities and
assistance. This study was supported by a grant
from the Netherlands Foundation for the
Advancement of Tropical Research (WOTRO),
˚berg, P. (1990). Measuring size and choosing category size for a
transition matrix study of the seaweed Ascophyllum nodosum.
Mar. Ecol. Prog. Ser., 63: 281–288.
˚berg,P.&Pav i a , H. (1997). Temporal and multiple scale spatial
variation in juvenile and adult abundance of the brown alga
Ascophyllum nodosum.Mar. Ecol. Prog. Ser., 158: 111–119.
Ang, P.O. (1991). Natural dynamics of a Fucus distichus
(Phaeophyceae, Fucales) population. Mar. Ecol. Prog. Ser., 78:
Ang, P.O. (2000). Relative importance of recruitment versus
regeneration in maintaining the populations of Sargassum
henslowianum in Ping Chau, Hong Kong. J. Phycol., 36: 2–3.
Bak, R.P.M. (1977). Coral reefs and their zonation in Netherlands
Antilles. Am. Ass. Petrol. Geol. USA Stud. Geology., 4: 3–16.
Blanchette, C.A. (1997). Size and survival of intertidal plants in
response to wave action: a case study with Fucus gardneri.
Coleman, M.A. (2002). Small-scale spatial variability in intertidal
and subtidal turfing algal assemblages and the temporal
generality of these patterns. J. Exp. Mar. Biol. Ecol., 267: 53–74.
Creed, J.C., Norton, T.A. & Harding, S.P. (1996a). The
development of size structure in a young Fucus serratus
population. Eur. J. Phycol., 31: 203–209.
Creed, J.C., Norton, T.A. & Kain, J.M. (1996b). Are neighbours
harmful or helpful in Fucus vesiculosus populations? Mar. Ecol.
Prog. Ser., 133: 191–201.
DeHaan,D.&Zaneveld, J.S. (1959). Some notes on tides in
Annabaai harbour, Curac¸ ao, Netherlands Antilles. Bull. Mar.
Sci. Gulf Caribb., 9: 224–236.
DePaula, E.J. & Oli veira Filho, E.C. (1982). Wave exposure
and ecotypical differentiation in Sargassum cymosum
(Phaeophyta-Fucales). Phycologia,21: 145–153.
DeRuyter van St eve ninck , E.D. & Breeman, A.M. (1987).
Population dynamics of a tropical intertidal and deep-water
population of Sargassum polyceratium (Phaeophyta). Aquatic
Bot., 29: 139–156.
Deysher, L.E. (1984). Reproductive phenology of newly intro-
duced populations of the brown alga, Sargassum muticum
(Yendo) Fensholt. Hydrobiologia,116/117: 403–407
Diaz-Piferrer, M. (1969). Distribution of the marine benthic
flora of the Caribbean Sea. Carib. J. Sci., 9: 151–178.
Engelen, A.H. (2004). Flexibility without compromise. Population
biology of the brown seaweed Sargassum polyceratium. PhD
thesis University of Groningen. ISBN 90-9017581–4.
Engelen, A.H., Olsen, J.L., Breeman, A.M. & Stam , W.T.
(2001). Genetic differentiation in Sargassum polyceratium
(Fucales, Phaeophyceae) around the island of Curac¸ ao
(Netherlands Antilles). Mar. Biol., 139: 267–277.
Espi noza, J. (1990). The southern limit of Sargassum muticum
(Yendo) Fensholt (Phaeophyta, Fucales) in the Mexican Pacific.
Bot.Mar., 33: 193–196.
Gay l o r d , B. (1999). Detailing agents of physical disturbance:
wave-induced velocities and accelerations on a rocky shore.
J.Exp.Mar.Biol.Ecol., 239: 85–124.
Hoffman n, A.J. & Santelices, B. (1991). Banks of
algal microscopic forms: hypotheses on their functioning
and comparisons with seed banks. Mar. Ecol. Progr. Ser.,
Hurd, C.L. (2000). Water motion, marine algal physiology and
production. J.Phycol., 36: 453–472.
Hurtado , A.Q. & Ragaza, A.R. (1999). Sargassum studies in
Currimao, Ilocos Norte, Northern Philippines I. Seasonal
variations in the biomass of Sargassum carpophyllum J.
Agardh, Sargassum ilicifolium (Turner) C. Agardh and
Sargassum siliquosum J. Agardh (Phaeophyta, Sargassaceae).
Bot.Mar., 42: 321–325.
Hwa n g , E.K. & Dring, M.J. (2002). Quantitative photoperiodic
control of erect thallus production in Sargassum muticum.
Bot.Mar., 45: 471–475.
Kareiva, O. (1990). Population dynamics in spatially complex
environments: theory and data. Phil.Trans.R.Soc.Lond.B,330:
Kendrick, G.A. & Walker, D.I. (1991). Dispersal distances for
propagules of Sargassum spinuligerum (Sargassaceae,
Phaeophyta) measured directly by vital staining and venturi
suction sampling. Mar.Ecol.Prog.Ser., 79: 133–138.
Kendrick, G.A. & Walker, D.I. (1994). Role of recruitment in
structuring beds of Sargassum spp. (Phaeophyta) at Rottnest
Island, Western Australia. J.Phycol.30: 200–208
Kendrick, G.A. & Walker, D.I. (1995). Dispersal of propagules
of Sargassum spp. (Sargassaceae: Phaeophyta): Observations
of local patterns of dispersal and consequences for recruit-
ment and population structure. J.Exp.Mar.Biol.Ecol., 192:
Levin, S.A. (1992). The problem of pattern and scale in ecology.
McCourt, R.M. (1984). Seasonal patterns of abundance,
distributions, and phenology in relation to growth strate-
gies of three Sargassum species. J.Exp.Mar.Biol.Ecol., 74:
Norton, T.A. (1991). Conflicting constraints on the form of
intertidal algae. British Phycol.J., 26: 203–218.
Norton, T.A., Math i e s o n, A.C. & Neus hul, M. (1981).
Morphology and environment. In: The Biology of Seaweeds.
(Lobban, C.S. & Wynne, M.J., editors), 421–451. University of
California Press, Berkeley.
Norton, T.A. & Deysher, L.E. (1989). The reproductive ecology
of Sargassum muticum at different latitudes. In: Reproduction,
Genetics,and Distributions of Marine Organisms. (Ryland, J.S. &
Tyler, P.A., editors), 147–152. Proc. 23rd European Mar. Biol.
Symp., Olsen and Olsen, Fredensborg, Denmark.
Prince, J.S. (1980). The ecology of Sargassum pteropleuron
Grunow (Phaeophyceae, Fucales) in the waters off South
Florida II. Seasonal photosynthesis and respiration of
S. pteropleuron and comparison of its phenology with that of
S. polyceratium Montagne. Phycologia,19: 190–193.
Sjtun ,K.&Fredriksen, S. (1995). Growth allocation in
Laminaria hyperborea (Laminariales, Phaeophyceae) in
relation to age and wave exposure. Mar.Ecol.Prog.Ser., 126:
Stiger,V.&Pay r i , C.E. (1999). Spatial and temporal patterns of
settlement of the brown macroalga Turbinaria ornata and
Sargassum mangarevense in a coral reef on Tahiti. Mar.Ecol.
Prog.Ser., 191: 91–100.
Tay l o r , W.R. (1960). Marine algae of the eastern tropical and
subtropical coasts of the Americas. University of Michigan Press,
Trono , G.C.J.R. & Tolentino, G.L. (1993). The reproductive
phenology of Sargassum spp. (Fucales, Phaeophyta) in Bolinao,
Pangasinan. Proc.2nd RP-USA Phycol.Sympos.Workshop 2:
Underwood, A.J. (1997). Experiments in Ecology: their logical
design and interpretation using analysis of variance. Cambridge
University Press, Cambridge.
Underwood, A.J. & Chapman, M.G. (1998). Variation in algal
assemblages on wave-exposed rocky shores in New South Wales.
Mar.Freshwater Res., 49: 241–254.
Van den Hoek, C. (1969). Algal vegetation-types along the open
coasts of Curacao, Netherlands Antilles. Proc.Kon.Ned.Akad.
Wet.Ser., 72: 537–577.
Effects of wave exposure and depth on Sargassum polyceratium 157
Van den Hoek, C., Cortel-Breeman, A.M. & Wanders, J.B.W.
(1975). Algal zonation in the fringing coral reef of Curac¸ ao,
Netherlands Antilles, in relation to zonation of corals and
gorgonians. Aquatic Bot., 1: 269–308.
Van den Hoek, C., Breeman, A.M., Bak, R.P.M. & Van Buurt,
G. (1978). The distribution of algae, corals and gorgonians in
relation to depth, light attenuation, water movement and grazing
pressure in the fringing coral reef of Curacao, Netherlands
Antilles. Aquatic Bot., 5: 1–46.
Van Duyl, F.C. (1985). Atlas of the living reefs of Curac¸ ao and
Bonaire (Netherlands Antilles). Foundation for Scientific
Research in Surinam and The Netherlands Antilles, Vol. 117,
Utrecht, The Netherlands.
Van Loenhoud, P.J. & Van de Sande, J.C.P.M. (1977). Rocky
shore zonation in Aruba and Curac¸ ao (Netherlands Antilles),
with the introduction of a new general scheme of zonation. Proc.
Kon.Ned.Acad.Wet.Ser. C., 80: 437–474.
Vermeij, M.J.A. (1997). Population ecology of Sargassum poly-
ceratium. MSc thesis, University of Groningen, The Netherlands,
Wanders, J.B.W. (1976). The role of benthic algae in the shallow
reef of Curac¸ ao (Netherlands Antilles) II: Primary productivity
of the Sargassum beds on the North-East coast submarine
plateau. Aquatic Bot., 2: 327–335.
Wheeler, W.N. (1980). Effect of boundary layer transport on the
fixation of carbon by the giant kelp Macrocystis pyrifera.Mar.
Biol., 56: 103–110.
Wheeler, W.N. (1988). Algal productivity and hydrodynamics – a
synthesis. Progr.Phycol.Res., 6: 23–58.
Winer, B.J., Brown, D.R. & Mic hels, K.M. (1991). Statistical
principles in experimental design. McGraw-Hill, New York.
Yoshida, T. (1983). Japanese species of Sargassum subgenus
Bactrophycus (Phaeophyta, Fucales). J.Fac.Sci.Hokkaido Univ.
Ser., 5(13): 99–246.
A. H. Engelen et al. 158