Content uploaded by Arthur R. Bos
Author content
All content in this area was uploaded by Arthur R. Bos
Content may be subject to copyright.
Content uploaded by Arthur R. Bos
Author content
All content in this area was uploaded by Arthur R. Bos
Content may be subject to copyright.
Mar Biol (2008) 156:55–63
DOI 10.1007/s00227-008-1064-2
123
ORIGINAL PAPER
Population dynamics, reproduction and growth
of the Indo-PaciWc horned sea star, Protoreaster nodosus
(Echinodermata; Asteroidea)
Arthur R. Bos · Girley S. Gumanao ·
Joan C. E. Alipoyo · Lemuel T. Cardona
Received: 10 March 2008 / Accepted: 16 September 2008 / Published online: 8 October 2008
© The Author(s) 2008. This article is published with open access at Springerlink.com
Abstract The horned sea star (Protoreaster nodosus) is
relatively common in the Indo-PaciWc region, but there is
little information about its biology. This study of the popu-
lation biology of P. nodosus was carried out in Davao Gulf,
The Philippines (7°5⬘N, 125°45⬘E) between September
2006 and May 2008. Protoreaster nodosus was found in
sand and seagrass dominated habitats at a mean density of
29 specimens per 100 m2 and a mean biomass of 7.4 kg per
100 m2, whereas a signiWcantly lower density and biomass
was found in coral and rock dominated habitats. Adult
specimens (mean radius R= 10.0 cm) were found at depths
of 0–37 m, whereas juveniles (R< 8 cm) were only found
in shallow sandy habitats with abundant seagrass (water
depth ·2 m). Increased gonad weights were found from
March to May (spawning period), which coincided with an
increasing water temperature and a decreasing salinity.
Density and biomass did not change signiWcantly during
reproduction, but sea stars avoided intertidal habitats. All
specimens with R> 8 cm had well developed gonads and
their sex ratio was 1:1. Protoreaster nodosus grew rela-
tively slowly in an enclosure as described by the exponen-
tial function G= 7.433 e¡0.257 £R. Maturing specimens
(R= 6–8 cm) were estimated to have an age of 2–3 years.
Specimens with a radius of 10 cm (population mean) were
calculated to have an age of 5–6 years, while the maximum
age (R= 14 cm) was estimated as 17 years. Potential eVects
of ornamental collection on the sea star populations are
discussed.
Introduction
Sea stars are common residents in coral reefs and most of
the Indo-PaciWc species are widely distributed in this region
(Clark and Rowe 1971 ; Colin and Arneson 1995). Despite
of, the commonness and abundance of some of these tropi-
cal Indo-PaciWc sea stars, there are remarkably few studies
of their biology and ecology. A reason for the lack of bio-
logical information seems to be that these sea stars have
never been harvested for consumption and thus have not
been of direct economic interest. With limited Wnancial
resources for research, Indo-PaciWc countries may prioritize
conducting biological studies about species or processes
that threaten economic interests. An example of such spe-
cies is the crown-of-thorns sea star (Acanthaster planci),
which has caused severe damage to large areas of coral
reefs (e.g. Moran et al. 1988).
The horned sea star (Protoreaster nodosus) is a rela-
tively large Indo-PaciWc asteroid which is commonly
Communicated by J.P. Grassle.
Electronic supplementary material The online version of this
article (doi:10.1007/s00227-008-1064-2) contains supplementary
material, which is available to authorized users.
A. R. Bos (&) · G. S. Gumanao · J. C. E. Alipoyo · L. T. Cardona
Research OYce, Davao del Norte State College,
New Visayas, 8105 Panabo City, The Philippines
e-mail: arthurrbos@yahoo.com
A. R. Bos
German Development Service, DED,
11th Floor PDCP Bank Center Building,
VA RuWno corner LP Leviste Streets,
Salcedo Village, 1227 Makati, The Philippines
A. R. Bos
National Museum of Natural History Naturalis,
P.O. Box 9517, 2300 RA Leiden, The Netherlands
A. R. Bos
Department of Environmental Science, Radboud University,
P.O. Box 9010, 6500 GL Nijmegen, The Netherlands
56 Mar Biol (2008) 156:55–63
123
encountered in relatively shallow waters. It primarily
inhabits seagrass meadows in the triangle formed by the
Seychelles, New Caledonia, and southern Japan. Adult
specimens reach a radius of approximately 15 cm
(Schoppe 2000) and generally feed on meiobenthos and
microbial/microalgal Wlms on sediment and seagrasses
(Thomassin 1976; Klumpp et al. 1993; Scheibling and
Metaxas 2008). Schoppe (2000) mentioned that P. nodosus
additionally preys on gastropods, sponges, soft corals, and
other small invertebrates. Scheibling and Metaxas (2008)
obtained baseline information on abundance and biomass
of P. nodosus in undisturbed habitats in Palau. They fur-
ther observed synchronous spawning under laboratory
conditions at full moon in May, but no data are available
on the spawning activity and behavior of this species from
other areas.
Sea stars are, among other invertebrates and reef Wsh
(Shuman et al. 2005), target species for the ornamental
trade (both live and dried) and there is a pressing need for
basic biological information and life history characteristics
of these organisms (Wabnitz et al. 2003). The usually
uncontrolled harvest of sea stars for ornamental purposes
has caused some species to completely disappear from
Wshed areas (personal observation in Cebu Island, Philip-
pines) and may have happened to a Caribbean asteroid
(Scheibling and Metaxas 2008). In general, the current
status of biological knowledge of sea stars needs to be
extended to better understand their population processes.
The present study describes the population dynamics, habi-
tat association, reproduction, and growth of the horned sea
star Protoreaster nodosus.
Materials and methods
Transects and environmental parameters
A total of 30 belt transects were used to study the habitat
association of Protoreaster nodosus along the western
coasts of Samal Island and Talikud Island in the Davao
Gulf, The Philippines (Fig. 1): 11 transects between Sep-
tember and December 2006, 12 in January 2007, and 7 in
March 2007. No more than two transects were done on a
single day. Transects were chosen to cover diVerent habitat
types (dominated by corals, limestone rock, sand, or
seagrasses) and at various depths. Each belt transect
(50 £5 m) was randomly positioned in a homogenous hab-
itat type and oriented parallel to the coastline at a constant
depth. Depth range of the transects was 0–33 m. Intertidal
and shallow habitats (<2 m depth) were surveyed using
snorkeling gear, whereas for deeper habitats SCUBA
equipment was employed.
Corals, crevices, and rocky structures within a belt tran-
sect were carefully searched, after which all encountered
sea stars were measured. Habitat composition was studied
every 10 m along each belt transect with a 1-m2 quadrat,
subdivided into 16 equal squares of 625 cm2. The dominat-
ing sediment type or the structural component of each
square determined the habitat type to which it was attrib-
uted. Subsequently, the contribution of each habitat cate-
gory was calculated for a whole belt transect. All transects
were carried out during the daytime, because a few night
observations did not result in changed numbers or size of
our target species.
Fig. 1 Map of Republic of the
Philippines with Davao Gulf
region magniWed on right.
Davao Gulf with Samal and
Talikud Islands, two major
cities, and the research station
(RS). White dots along western
coasts of the islands indicate
locations of transects (n=30).
Isobaths at 300 m intervals
Mar Biol (2008) 156:55–63 57
123
Water temperature and salinity were measured with a
thermometer (0.5°C accuracy) and refractometer (1 U accu-
racy) from water collected at 3-m depth during every visit
to the research station in Samal Island (Fig. 1).
To study spawning-related changes in sea star density,
three Wxed belt transects (50 £5 m) were studied on, or
one day prior to, every full moon from September 2006 to
September 2007. These transects, diVerent from those used
for the habitat study, were chosen to cover the natural
habitats of Protoreaster nodosus at an intertidal, a shallow
subtidal, and a relatively deeper subtidal location (Sl, Sup-
plementary materials). The tidal habitat was dominated by
limestone rock and bare sand, whereas the subtidal habitats
were dominated by bare sand and seagrass, mainly Enhalus
acoroides.
Size measurements
The size of a sea star was determined by calculating a mean
radius R (=ray or arm length) for rays A and B, following
the Carpenter orientation (Carpenter 1884). Radius was
measured from the mouth (center of the disk on the oral
side) along the ambulacral groove to the tip of a ray using
calipers (1 mm accuracy).
A radius weight relationship was determined to calcu-
late sea star biomass. To accomplish that, intact penta-
symmetrical specimens were collected and taken to the
research station. Radius was determined as described
above, but in this case all Wve rays were measured. Wet
weight was determined with a portable balance (1 g
accuracy) after removing the coelomic Xuid by gently
squeezing the specimen so as to increase weighing accu-
racy. A small incision in the body wall of one ray helped
to further drain the coelomic Xuid from larger specimens
(Scheibling 1980a). Immediately after weighing, sea
stars were returned to the location of collection. The
following signiWcant relationship between radius and
weight was found: W= 0.0028 R2.4743 (n= 78, r= 0.991;
Bos et al. 2008b) and was used to calculate sea star
weight and biomass for all specimens counted in tran-
sects.
Gonad analysis
To study the gonad development and to determine the tim-
ing of spawning of Protoreaster nodosus, 12 specimens
were collected one day prior to or on the day of every full
moon between November 2006 and October 2007. Speci-
mens collected for gonad dissection had a radius of 6–
13 cm; smaller specimens were only used to determine at
what size P. nodosus Wrst reproduces. After measuring (Wve
rays) and weighing, the specimens were stored in a freezer
until dissection. Due to extended power cuts, samples were
partially and fully deteriorated in February and June 2007,
respectively.
The gonads were dissected from all Wve rays, weighed
(0.001 g accuracy) and sexed according to color: female
gonads are orange/yellow, male gonads are creamy white
(described for other species by Scheibling 1981a). The
gonado-somatic index (GSI) was calculated as GSI = (GW/
BW) £100, where GW is the gonad wet weight (g) and
BW is the drained body wet weight (g). Furthermore, the
gonads and soma were dried at 105°C for 24 h to determine
dry weight and water content.
Growth experiment
To study the growth rate of P. nodosus, 53 sea stars with
initial radii of 3.2–7.8 cm were kept in an enclosure from
September 2006 to May 2008. The enclosure (25 m2 and
1.2 m high), positioned at a water depth of 0.5 m during
low tide, was made of dense netting material for its sides
(mesh size 1 cm) and Wsh net on the top (mesh size 3 cm).
The densities of seagrasses, macroalgae, and corals inside
the enclosure were similar to those outside the enclosure,
where P. nodosus was abundant. P. nodosus was not fed
during the experiment, because abundant food [microalgal
Wlms, meiobenthos (Scheibling and Metaxas 2008), and
small invertebrates] was available in the enclosure. Sea star
density inside the enclosure was kept lower than high den-
sities observed in adjacent natural habitats to avoid density
dependent eVects. Sea stars were measured (Wve rays) and
weighed once every 2 months. Numerals were inscribed on
an inter-radial area on the oral side or in an aboral inter-ray
with a sharp pencil to individually discriminate the sea
stars. Markings were legible for about 2 months and
renewed if needed. This procedure has no adverse eVect on
individuals (Scheibling 1980b). Photographs additionally
helped to recognize individuals when markings had faded.
Some sea stars escaped and new ones entered the enclosure
when the nets were damaged by storms. The measurements
of those were excluded from the data set. Mortality was not
observed in the enclosure. Growth was considered not to be
aVected by seasons.
Data analysis
DiVerences in sea star density and biomass between habitat
types were tested using one-way ANOVA. The Levene’s
test was used to test for homogeneity of variances; data
were log-transformed in case of heterogeneity. Post-hoc
analysis was done with the Tukey’s honestly signiWcant
diVerence test.
A small range of sea star sizes was selected for the GSI
comparison between sexes to avoid size dependent eVects.
The radius range for which the GSI was >1 was selected to
58 Mar Biol (2008) 156:55–63
123
assure that juveniles were excluded from the analysis.
Similarly, a few large specimens were excluded. GSI diVer-
ences between sexes were tested with the Student’s t test.
Although every full moon 12 specimens were collected, the
sex of a few could not be determined. Therefore, GSI-val-
ues were pooled for two succeeding months to increase the
sample sizes. Chi-square (2) analysis was used to test the
sex distribution. DiVerences between sexes were tested
with the Student’s t test.
To calculate the growth rate in the enclosure, sea stars
were categorized into 0.5 cm groups, based on their mean
radius at stocking. Individual observations were restricted
to a maximum period of 7 months (=four observations per
sea star) so as to allow a linear function to describe growth
in time. Linear regressions were performed for each avail-
able size group. Subsequently, the slope of each linear
function was used to calculate an exponential relationship
between sea star growth and radius.
Results
Habitat association
In total 1,059 specimens of the horned sea star, Protoreaster
nodosus, were found in 20 transects in diVerent habitats
(Fig. 2). A Mean density of 23 and 34 specimens per
100 m2 was, respectively found in sand and seagrass domi-
nated habitats, whereas lower numbers were found in coral
and rock dominated habitats (Fig. 2a). Densities of >70
specimens per 100 m2 were only found in seagrass
dominated habitats. Sea star density was found to be
signiWcantly diVerent between habitat types (ANOVA,
log-transformed data: F3,19 = 5.361, P= 0.010). Sea star
densities in both sand and seagrass dominated habitats were
signiWcantly higher than in coral dominated habitats
(Tukey’s test, P< 0.05). A mean sea star biomass of 8.5
and 6.3 kg per 100 m2 was found in sand and seagrass dom-
inated habitats, respectively (Fig. 2b), and the biomass was
signiWcantly diVerent between habitat types (ANOVA, log-
transformed data: F3,19 = 3.665, P= 0.035). A post-hoc
comparison showed that biomass diVered signiWcantly
between coral and sand dominated habitats (Tukey’s test,
P< 0.05). A maximum biomass of 18.6 kg per 100 m2 and
three other observations of >10 kg per 100 m2 were found
in habitats dominated by sand and seagrass.
Large specimens were found between 0 and 33 m depth,
whereas small specimens were exclusively found in shal-
low habitats (Fig. 3). Specimens with radii ·7.9 cm were
never found in water >2 m with only one exception: a speci-
men of 7.2 cm at 10 m depth. Furthermore, all specimens
with radii ·5.5 cm were found in two transects with a sea-
grass cover of 94%. One specimen was found at a depth of
37 m in a sandy habitat outside the transects.
Fig. 2 Protoreaster nodosus. a Mean density (specimens per
100 m2)§SE and b mean biomass (kg per 100 m2)§SE in habitats
dominated by coral (n= 13), limestone rock (n= 3), sand (n=8), and
seagrass (n=6)
Fig. 3 Protoreaster nodosus. Mean radius §SD (cm) versus transect
depth (m) in 20 transects in Davao Gulf from September 2006 to
March 2007. P. nodosus was not found in 10 other transects
Mar Biol (2008) 156:55–63 59
123
Reproduction and related behavior
The GSI of specimens with radii of 6–13 cm ranged from
0 to 6.7 (S2, Supplementary material). Specimens with a
radius <8.0 cm had a GSI < 1, with two exceptions. Also,
all specimens with a radius >12.5 cm had a GSI < 2 (S2,
Supplementary material). To avoid size dependent eVects in
the GSI analysis, only specimens with radii of 8.0–12.5 cm
were considered.
The GSI for females was signiWcantly higher than for
males in the three consecutive periods November–Decem-
ber 2006 (t=2.64, df =16, P< 0.05), January–February
2007 (t=2.32, df =9, P< 0.05) and March–April 2007
(t=3.05, df =19, P<0.01; Fig.4a). For females, the mean
GSI reached its maximum of 4.5 in March 2007 and con-
tinued to be relatively high until May 2007. For males, the
mean GSI reached its maximum of >2 in April and May
2007. The period of high GSIs coincided with increasing
water temperatures, from 27.0 to 30.0°C, and decreasing
salinity, from 35 to 31 (Fig. 4b). The GSI was not signiW-
cantly diVerent between males and females in the periods
May–June (t=0.14, df =9, P> 0.05), July–August (t=1.31,
df = 18, P> 0.05) and September–October 2007 (t=0.61,
df = 11, P> 0.05). Ripe adults of both sexes, but especially
males, were found from March to August 2007 with an
increasing number of spent adults towards August.
Specimens with radii of 8.0–12.5 cm had a male/female
ratio of 1.2 (Table 1), which was not signiWcantly biased
towards either sex (2 test, P> 0.05). Five percent of the
specimens (R< 10.4 cm) could not be sexed. No signiWcant
diVerences between males and females were found for
radius (t=1.34, df =92, P> 0.05), wet weight (t=1.41,
df = 89, P> 0.05), and water content (t=0.30, df =82,
P> 0.05; Table 1).
In total 3,092 specimens, Protoreaster nodosus were
found in three Wxed transects at diVerent depths during 14
full moon samplings from September 2006 to September
2007, and used to assess behavioral changes related to
reproduction (Fig. 5). The density of P. nodosus in the
intertidal transect [0 m at mean low tide (MLT)] generally
ranged from 5 to 20 specimens per 100 m2, but hardly any
specimen was found from March to June 2007. Similarly,
the density of P. nodosus in the shallow subtidal transect
(0.5 m at MLT) was remarkably low from February to May
2007, whereas the density ranged from 30 to 50 specimens
per 100 m2 during the rest of the observations. In the deeper
subtidal transect (2 m at MLT), the density of P. nodosus
steadily increased from about 40 specimens per 100 m2 in
November 2006 to a maximum of 82.4 specimens per
100 m2 in May 2007 (Fig. 5). After May the density
decreased again to about 40 specimens per 100 m2.
The biomass of P. nodosus followed a similar trend as
the density, because the size range of these sea stars was
relatively small (>95% within R= 8–12 cm). The mean
biomass of the intertidal, shallow subtidal, and deeper sub-
tidal transects were 3.1, 8.2, and 11.8 kg per 100 m2 respec-
tively. A maximum biomass of 20.8 kg per 100 m2 was
found in the deeper subtidal transect in May 2007.
Table 1 Protoreaster nodosus
Number and corresponding percentages of males, females, and specimens that could not be sexed with a radius range from 8.0 to 12.5 cm. Mean
radius, mean wet weight (g) and mean water content (%) are presented with the SD between brackets
Number of specimen Mean radius (cm) Mean wet weight (g) Mean water content (%)
Male 50 (52%) 10.3 (1.3) 281.3 (82.1) 52.7 (2.3)
Female 41 (43%) 10.6 (1.2) 304.7 (74.6) 53.1 (3.3)
Unsexed 5 (5%) 8.9 (0.9) – –
Fig. 4 Protoreaster nodosus. a Mean gonado-somatic index §SE for
females (Wlled circle) and males (inverted triangle) from November
2006 to October 2007. Each month, 12 specimens were collected and
the number of sexed males and females are presented b Water temper-
ature (°C, thin line) and salinity (bold line) for the same period
60 Mar Biol (2008) 156:55–63
123
The radius frequency distribution of P. nodosus was uni-
modal throughout the year with a relatively constant mean
radius of 10.0 (§1.0) cm (S3, Supplementary material).
However, the mode was 9.5 cm from September 2006 to
January 2007 and changed to 10.0 cm from February to
September 2007 (S3, Supplementary material). In Novem-
ber 2006 an increased number of smaller specimens
(R< 6 cm) were found.
Sea star growth
SigniWcant linear relationships between sea star radius
and time of observation were found for ten 0.5 cm size
groups in the growth experiment (S4, Supplementary
material). The slope of each linear relationship was used
to represent the growth rate of the respective size group.
Generally, growth rates decreased with increasing radius
(S4, Supplementary material). The highest growth rate
of 3.1 cm per year was observed for the sea stars with a
mean initial radius of 3.9 cm (Fig. 6). The largest sea
stars had a mean initial radius of 7.7 cm and grew at a
rate of 1.0 cm per year. An exponential function
described the relationship between sea star growth and
radius (G=7.433£e¡0.257 £R; Fig. 6). This function
was used to estimate the age of sea stars at a given size.
Assuming the age of a horned sea star with a radius of
3.0cm is 1year (see “Discussion”) and the correspond-
ing growth rate is 3.4 cm per year (Fig. 6), the radius at
age 2 would then become 6.4 cm (=3.0 + 3.4), which
corresponds with a growth rate of 1.4 cm per year
(Fig. 6). Continuing this iteration, maturing specimens
with a radius between 6.0 and 8.0 cm were estimated to
be 2–3 years. An individual with a radius of 10 cm was
calculated to have an age of 5–6 years, and the largest
specimen observed in the present study (R=14cm) was
estimated to be 17 years.
Discussion
Behavioral change during the reproductive period is com-
monly observed in echinoderms (Lamare and Stewart
1998; McCarthy and Young 2004). For example, spawning
aggregations have been observed among sea stars and
apparently increase reproductive success (Scheibling
1980b; Hamel and Mercier 1995; Metaxas et al. 2002). On
the other hand, Scheibling and Metaxas (2008) found Pro-
toreaster nodosus to be randomly distributed during the
spawning season and showed that, when experimentally
aggregated, they rapidly returned to the pre-manipulation
density. The relatively high densities and biomasses of
P. nodosus observed during the present study partially
occurred after the spawning season and thus were not
directly related to reproductive behavior. Also, Babcock
et al. (1994) observed only small changes in the density of
the crown-of-thorns sea star during spawning periods and
suggested that high levels of aggregation may not be
necessary, because this species performs successful long-
distance fertilizing. Moreover, Babcock et al. (1994)
predicted that migration to shallow water may have a mea-
surable eVect on the fertilization success of sea stars, while
we observed P. nodosus avoiding intertidal habitats during
the reproductive period; the risk of being excluded from
reproduction during low tide in intertidal habitats may
have triggered adults to move away from these areas.
Although local aggregations were not observed in P. nodo-
sus populations, migrating to certain water depths may be
a strategy to increase fertilization success.
Female gonads of P. nodosus were heavier than those of
males of similar size and, as a consequence, females had a
signiWcantly higher GSI than males. Generally, the gonad
volume of female sea stars is known to be larger than that
Fig. 5 Protoreaster nodosus. Density (specimens per 100 m2) in three
Wxed transects during full moon from September 2006 to September
2007. Intertidal transect (Wlled circle = 0 m at MLT), shallow subtidal
transect (open circle = 0.5 m at MLT), deeper subtidal transect (invert-
ed triangle = 2 m at MLT) Fig. 6 Protoreaster nodosus. Growth rate (cm per year) as a function
of the radius (cm) as observed for 53 specimens between September
2006 and May 2008. Drop lines represent age estimations
Mar Biol (2008) 156:55–63 61
123
of males (Scheibling 1981a). Males, however, were ripe for
a longer period than females. In fact, Scheibling (1981a)
found spawning to be relatively synchronous in several
Oreaster reticulatus populations, a tropical Atlantic sea star
similar to P. nodosus, assuming that spawning synchrony is
important for reproduction success.
High GSI-values for P. nodosus were found from March
to May, which coincided with increasing water tempera-
tures and decreasing salinity in the Davao Gulf. The advan-
tage of such timing may be that the best growth conditions
exist during the larval period and settlement. Many coral
species in the Philippines also spawn from March to June
(Vicentuan et al. 2007) and Scheibling and Metaxas (2008)
observed populations of P. nodosus in Palau spawning at
full moon in May under laboratory conditions. The length
of the larval period of P. nodosus is unknown, but informa-
tion about other tropical asteroids is available. Williams
and Benzie (1993) found the larval period of Linckia
laevigata and Acanthaster planci to be 28 and 14 days,
respectively, and Moran (1988) described the larval period
of A. planci to be 3 weeks. By comparison, Scheibling (per-
sonal communication) found the larval period of the tropi-
cal Atlantic sea star O. reticulatus to be 24 days. Assuming
the larval period of P. nodosus to have a maximum length
of 1 month, settlement would have taken place, at the latest,
in July. After settlement, a growth rate of 6.5 cm per year
(at R= 0.5 cm) could be expected based on the growth rates
found for larger juveniles (Fig. 6). By the end of the Wrst
year (6 months later), these new recruits may reach a radius
of about 3 cm. Therefore, the increased number of juveniles
with radii of about 5 cm found in November 2006 (S3, Sup-
plementary material) probably did not represent a new
cohort migrating to adult habitats. Migration from juvenile
to adult habitats may be a more gradual process that takes
place over a relatively long period. This process could also
explain why no juvenile cohorts were found entering the
adult population in the radius frequency distributions.
Juveniles of P. nodosus were found in extremely high
densities in shallow habitats only. One reason may be that
juveniles are associated with seagrass dominated habitats,
which are only found in shallow waters. Scheibling and
Metaxas (2008) also found high densities of juvenile
P. nodosus in seagrass beds and suggested that their darker
coloration, compared to adults, help to camouXage. Indeed,
juveniles seemed to prefer dense meadows of thin-leaved
seagrass species, such as Thalassia spp. (Bos et al. 2008b).
A second reason may be that juveniles avoid predation by
roaming in seagrass habitats. This seems in agreement with
Bos et al. (2008a) who observed the solitary corallimorph,
Pseudocorynactis sp., to feed on P. nodosus specimens in a
shallow patch reef. Similarly, Scheibling (1980a) suggested
that juveniles of Oreaster reticulatus were observed exclu-
sively in very dense seagrass meadows because they
provide greater protection from predators. Third, shallow
habitats may provide abundant food and favorable growth
conditions due to the relatively high water temperatures.
Several tropical sea stars have been observed at similarly
high densities and biomasses as P. nodosus. Scheibling and
Metaxas (2008) found P. nodosus at densities up to 70
specimens per 100 m2 and biomasses up to 20 kg per
100 m2. Also Oreaster reticulatus was found at a density
and biomass as high as 14.1 specimens per 100 m2 and
5.5 kg 100 m2 (Scheibling 1980b). Such dense populations
must, at least temporarily, have access to suYcient food.
Therefore, high densities of P. nodosus were expected to be
supported by abundant prey availability. P. nodosus is a
microphagous feeder (Scheibling and Metaxas 2008) and
furthermore preys on gastropods, sponges, soft corals, and
other small invertebrates (Schoppe 2000). Heart urchins,
present in high densities in most of the surveyed shallow
habitats, were often found to be digested by P. nodosus
(personal observation), which was also reported for
O. reticulatus (Scheibling 1982). The opportunistic feeding
strategy of P. nodosus may support the dense populations
as observed in the present study.
Growth in asteroids depends on food availability (quan-
tity and quality) and populations may suVer from density
dependent eVects (Feder and Christensen 1966). Accord-
ingly, Feder and Christensen (1966) concluded that size-
frequency distributions more adequately reXect the trophic
environment than the age structure of a population. The
unimodal size-frequency distributions found in natural hab-
itats during the present study (S3, Supplementary material)
did not allow accurate age estimations. However, we deter-
mined age by studying growth in an enclosure. The radius
and age at maturation was 6.0–8.0 cm and 2–3 years and
specimens with a mean radius of 10.0 cm were estimated to
have an age of 5–6 years (Fig. 6). Scheibling and Metaxas
(2008) suggested, maturation of P. nodosus to be reached at
a radius of about 11 cm, which was based on size and color-
ation. However, they found a maximum radius of 19.5 cm,
compared to a maximum radius of 14.0 cm in the present
study. Density dependent eVects however, seem not to
explain the observed size diVerence between populations in
the Davao Gulf and Palau, because Scheibling and Metaxas
(2008) found similar sea star densities. The role of intra- or
inter-speciWc food competition needs to be further studied.
Scheibling (1980b) found that the tropical Atlantic sea
star Oreaster reticulatus with a radius of 11–12 cm feeding
on natural sediments grew <1 cm per year. Moreover, no
net growth was observed for specimens with a radius of
13 cm. Similarly, growth of P. nodosus was <1 cm per year
for specimens with a radius >8 cm in the growth experi-
ment (Fig. 6). Also, the mode of the radius in natural popu-
lations increased from 9.5 to 10 cm during a 1-year period
(S3, Supplementary material). Although Scheibling and
62 Mar Biol (2008) 156:55–63
123
Metaxas (2008) found P. nodosus to have a similar size
range as the Atlantic sea star O. reticulatus, the specimens
studied in the Davao Gulf were smaller. Therefore, diVer-
ences in growth may be explained by the larger maximum
size: O. reticulatus reaches maturity at a radius of 12 cm
(Scheibling 1981a), mean reproductive size at a radius of
15 cm (Guzmán and Guevara 2002), and a maximum radius
of 21 cm (Scheibling 1981b).
The sea star Protoreaster nodosus, a target species in the
ornamental trade, was found to be a relatively slow grow-
ing species. Exploitation of slow growing species needs
careful consideration, because natural replenishment of
adult stocks is low and over-Wshing can easily occur. More-
over, specimens collected for trade usually have a radius of
<5 cm, which means that the trade of P. nodosus mainly
focuses on juveniles (personal observation). The present
study provides a basis for understanding the population
dynamics of P. nodosus and could be used to develop man-
agement plans in those areas where this species has been
collected. However, diVerences between populations (such
as observed in the Davao Gulf and Palau) underline the
need for monitoring local populations that are being
harvested.
Acknowledgments We greatly acknowledge E. Santos and K.
Schröder for supporting the initiation of the project. We thank I. Ebol,
C. Ganadores, E. Glimada, J. Lagarteja, B. Müller, S. Nitza, S. A.
Nitza, D. Padrogane, C. Petiluna, M. Saceda, F. Salac, J. Salinas, I.
Santamaria and R. Tejada for supporting the Weldwork. Furthermore,
we would like to thank R. Scheibling, B. Wilkinson, and two anony-
mous reviewers for their comments on an earlier version of this manu-
script. We for Wne-tuning the English language. We are grateful to the
communities from Samal Island and Talikud Island for sharing their
knowledge of the local marine resources. The performed experiments
complied with the current laws of the Republic of the Philippines.
Open Access This article is distributed under the terms of the Crea-
tive Commons Attribution Noncommercial License which permits any
noncommercial use, distribution, and reproduction in any medium,
provided the original author(s) and source are credited.
References
Babcock RC, Mundy C N, Whitehead D (1994) Sperm diVusion models
and in situ conWrmation of long-distance fertilization in the free-
spawning asteroid Acanthaster planci. Biol Bull 186:17–28.
doi:10.2307/1542033
Bos AR, Gumanao GS, Salac FN (2008a) A newly discovered predator
of the crown-of-thorns starWsh. Coral Reefs 27:581. doi:10.1007/
s00338-008-0364-9
Bos AR, Alipoyo JCE, Cardona LT, Gumanao GS, Salac FN (2008b)
Population structure of common Indo-PaciWc sea stars in the
Davao Gulf, Philippines. In: Proceedings of 9th biannual meeting
philippine association of marine science. UPV J Nat Sci 13
(accepted)
Carpenter PH (1884) Report upon the Crinoidea collected during the
voyage of HMS Challenger during the years 1873–1876. Part I:
General morphology, with descriptions of the stalked crinoids.
Rept ScientiWc Results HMS Challenger (s: Zool.) 11:1–422
Clark AM, Rowe FEW (1971) Monograph of shallow-water Indo-
West PaciWc echinoderms No. 690. British Museum (Nat. Hist.),
London
Colin PL, Arneson C (1995) Tropical paciWc invertebrates. Coral Reef
Press, California
Feder HM, Christensen AM (1966) Aspects of asteroid biology. In:
Boolootian R (ed) Physiology of echinodermata. Wiley, New
York, pp 87–127
Guzmán HM, Guevara CA (2002) Annual reproductive cycle, spatial
distribution, abundance and size structure of Oreaster reticulatus
(Echinodermata: Asteroidea) in Bocas del Toro, Panama. Mar
Biol (Berl) 141:1077–1084. doi:10.1007/s00227-002-0898-2
Hamel JF, Mercier A (1995) Prespawning behavior, spawning, and
development of the brooding starWsh Leptasterias polaris. Biol
Bull 188:32–45. doi:10.2307/1542065
Klumpp DW, Salita-Espinosa JT, Fortes MD (1993) Feeding ecology
and trophic role of sea urchins in a tropical seagrass community.
Aquat Bot 45:205–229. doi:10.1016/0304-3770(93)90022-O
Lamare MD, Stewart BG (1998) Mass spawning by the sea urchin
Evechinus chloroticus (Echinodermata: Echinoidea) in a New
Zealand Word. Mar Biol (Berl) 132:35–140. doi:10.1007/
s002270050379
McCarthy DA, Young CM (2004) EVect of water-borne gametes on
the aggregation behavior of Lytechinus variegatus. Mar Ecol Prog
Ser 283:191–198. doi:10.3354/meps283191
Metaxas A, Scheibling RE, Young CM (2002) Estimating fertilization
success in marine benthic invertebrates: a case study with the
tropical sea star Oreaster reticulatus. Mar Ecol Prog Ser 226:87–
101. doi:10.3354/meps226087
Moran PJ (1988) Crown-of-thorns starWsh: questions and answers.
Australian Institute of Marine Science, Townsville
Moran PJ, Bradbury RH, Reichelt RE (1988) Distribution of recent
outbreaks of the crown-of-thorns starWsh (Acanthaster planci)
along the Great Barrier Reef, 1985–1986. Coral Reefs 7:125–137.
doi:10.1007/BF00300972
Scheibling RE (1980a) Abundance, spatial distribution and size struc-
ture of populations of Oreaster reticulatus (Echinodermata:
Asteroidea) in seagrass beds. Mar Biol (Berl) 57:95–105.
doi:10.1007/BF00387375
Scheibling RE (1980b) Abundance, spatial distribution and size struc-
ture of populations of Oreaster reticulatus (Echinodermata:
Asteroidea) on sand bottoms. Mar Biol (Berl) 57:107–119.
doi:10.1007/BF00387376
Scheibling RE (1981a) The annual reproductive cycle of Oreaster
reticulatus (L.) (Echinodermata: Asteroidea) and interpopulation
diVerences in reproductive capacity. J Exp Mar Biol Ecol 54:39–
54. doi:10.1016/0022-0981(81)90101-5
Scheibling RE (1981b) Growth and respiration rate of juvenile Oreast-
er reticulatus (L.) (Echinodermata: Asteroidea) on Wsh and algal
diets. Comp Biochem Physiol 69A:175–176
Scheibling RE (1982) Habitat utilization and bioturbation by Oreaster
reticulatus (Asteroidea) and Meoma ventricosa (Echinoidea:
Spantagoidea) in a subtidal sand patch habitat. Bull Mar Sci
32:624–629
Scheibling RE, Metaxas A (2008) Abundance, spatial distribution, and
size structure of the sea star Protoreaster nodosus in Palau, with
notes on feeding and reproduction. Bull Mar Sci 82:211–235
Schoppe S (2000) Echinoderms of the Philippines. Times Edition,
Singapore
Shuman CS, Hodgson G, Ambrose RF (2005) Population impacts of
collecting sea anemones and anemoneWsh for the marine
aquarium trade in the Philippines. Coral Reefs 24:564–573.
doi:10.1007/s00338-005-0027-z
Mar Biol (2008) 156:55–63 63
123
Thomassin BA (1976) Feeding behaviour of the felt-, sponge-, and
coral-feeder sea stars, mainly Culcita schmideliana. Helgol Wiss
Meeresunters 28:51–65. doi:10.1007/BF01610796
Vicentuan KC, Guest JR, Baria MV, Cabaitan PC, Dizon RM,
Villanueva RD, Aliño PM, Edwards AJ, Gomez ED, Heyward AJ
(2007) Multi-species spawning of corals in north-western Philip-
pines. Coral Reefs 27:83. doi:10.1007/s00338-007-0325-8
Wabnitz C, Taylor M, Green E, Razak T (2003) From ocean to aquar-
ium. UNEP-WCMC, Cambridge, UK
Williams ST, Benzie JAH (1993) Genetic consequences of long larval
life in the starWsh Linckia laevigata (Echinodermata: Asteroidea)
on the Great Barrier Reef. Mar Biol (Berl) 117:71–77.
doi:10.1007/BF00346427
