Post-Settlement Life Cycle Migration Patterns and Habitat Preferences of Coral Reef Fish that Use Seagrass and Mangrove Habitats as Nurseries (Cocheret de la Moriniere et al, 2002)
Estuarine, Coastal and Shelf Science (2002) 55, 309–321
doi:10.1006/ecss.2001.0907, available online at http://www.idealibrary.com on
Post-settlement Life Cycle Migration Patterns and
Habitat Preference of Coral Reef Fish that use
Seagrass and Mangrove Habitats as Nurseries
E. Cocheret de la Morinierea, B. J. A. Polluxa, I. Nagelkerkena and
`
G. van der Veldea,b
a
Department of Animal Ecology and Ecophysiology, University of Nijmegen, Toernooiveld 1, 6525 ED Nijmegen,
The Netherlands
Received 14 May 2001 and accepted in revised form 8 October 2001
Mangroves and seagrass beds have received considerable attention as nurseries for reef fish, but comparisons have often
been made with different methodologies. Thus, relative importance of different habitats to specific size-classes of reef fish
species remains unclear. In this study, 35 transects in 11 sites of mangroves, seagrass beds and coral reef were surveyed
daily, in and in front of a marine bay on the island of Curacao (Netherlands Antilles). The density and size-frequency of
¸
nine reef fish species (including herbivores, zoobenthivores and piscivores) was determined during a five-month period
using a single methodology, viz. underwater visual census. All species were ‘ nursery species ’ in terms of their high
densities of juveniles in mangroves or seagrass beds. Relative density distribution of the size-classes of the selected species
over mangroves and seagrass beds suggested high levels of preference for either mangroves or seagrass beds of some
species, while other species used both habitats as a nursery. Spatial size distribution of the nine species suggested three
possible models for Post-settlement Life Cycle Migrations (PLCM). Haemulon sciurus, Lutjanus griseus, L. apodus, and
Acanthurus chirurgus appear to settle and grow up in bay habitats such as mangroves and seagrass beds, and in a later stage
migrate to the coral reef (Long Distance PLCM). Juveniles of Acanthurus bahianus and Scarus taeniopterus were found only
in bay habitats at close proximity to the coral reef or on the reef itself, and their migration pattern concerns a limited
spatial scale (Short Distance PLCM). Some congeneric species carry out either Long Distance PLCM or Short Distance
PLCM, thereby temporarily alleviating competition in reef habitats. Haemulon flavolineatum, Ocyurus chrysurus and Scarus
iserti displayed a Stepwise PLCM pattern in which smallest juveniles dwell in the mouth of the bay, larger individuals then
move to habitats deeper into the bay, where they grow up to a (sub-) adult size at which they migrate to nearby coral reef
habitats. This type of stepwise migration in opposite directions, combined with different preference for either mangroves
or seagrass beds among (size-classes of) species, shows that reef fish using in-bay habitats during post-settlement life
stages may do so by choice and not merely because of stochastic dispersal of their larvae, and underline the necessity of
2002 Elsevier Science Ltd. All rights reserved.
these habitats to Caribbean coral reef systems.
Keywords: fish; nursery grounds; mangrove swamps; seagrass; coral reef; Caribbean Sea; migration; Curacao
¸
Introduction efficiency, lower predator abundance, and high inter-
ception rate of the vegetation to planktonic larvae
In many studies, juveniles of reef fish species were
(Odum & Heald, 1972; Carr & Adams, 1973; Ogden
found in high densities in non-reef habitats, while the
& Ziemann, 1977; Blaber & Blaber, 1980; Shulman,
adults were found almost exclusively on the coral
1985; Parrish, 1989; Heck & Crowder, 1991;
reef itself (Pollard, 1984; Parrish, 1989). From this
Robertson & Blaber, 1992).
spatially heterogeneous size-frequency distribution,
Most authors, however, have focused on one or
Post-settlement Life Cycle Migration (PLCM) pat-
two habitats of the mangrove-seagrass-reef con-
terns were suggested that gave birth to the nursery
tinuum, often with different sampling methods, thus
concept. Mangroves and seagrass beds are considered
complicating comparisons among studies and among
nurseries to some reef fish species in the Western
habitats (e.g. Robertson & Duke, 1987; Thayer et al.,
Atlantic, Indian Ocean and Pacific Ocean (Pollard,
1987; Yanez-Arancibia et al., 1988; Blaber et al.,
´˜
1984; Parrish, 1989). Several authors have suggested
1989; Baelde, 1990; Rooker & Dennis, 1991;
the benefits of nurseries to juvenile reef fish, vary-
Sedberry & Carter, 1993; Laegdsgaard & Johnson,
ing from high food availability to lower predation
2001). Quantitative data on ontogenetic shifts in
habitat use from nursery to adult reef association are
b
Corresponding author. E-mail: gerardv@sci.kun.nl
2002 Elsevier Science Ltd. All rights reserved.
0272–7714/01/080309+13 $35.00/0
310 E. Cocheret de la Moriniere et al.
`
largely lacking (Ogden & Ehrlich, 1977; Weinstein & Methods
Heck, 1979; Rooker & Dennis, 1991; Appeldoorn
et al., 1997; Nagelkerken et al., 2000a) and the relative Study area
importance of these nurseries to different size-classes
The present study was carried out in Spanish Water
of reef fish species is still poorly known (Ogden &
Bay in Curacao, Netherlands Antilles (Figure 1).
¸
Gladfelter, 1983; Birkeland, 1985).
This 3 km2 bay is shallow (largely <6 m deep), har-
In recent underwater visual surveys in seven
bours extensive seagrass meadows and is fringed by
different habitats in a marine island bay in Curacao ¸
mangroves (Rhizophora mangle). Water depths under
(Netherlands Antilles), a number of reef fish species of
mangrove canopies ranged between 0·8 m and 1·8 m.
which juveniles were highly abundant in bay environ-
These canopies provide dark habitats (average light
ments were identified and grouped as ‘ nursery
extinction underwater was 85%, as opposed to 40%
species ’ (Nagelkerken et al., 2000a). Nagelkerken
over seagrass beds). The seagrass beds are dominated
et al. (2000a) used a low frequency of surveys in a
by monospecific stands of Thalassia testudinum
large number of transects, and focused on fish com-
(Kuenen & Debrot, 1995). Mean shoot density
munity structure in a range of habitats (mangroves,
( SD) in the seagrass transects was 246 m2 ( 110)
seagrass beds, algal beds, channel, fossil reef terrace
and seagrass canopy height averaged 28·0 cm
notches, boulders, coral reef). Of all habitats in that
( 11·5).
study, seagrass meadows proved to contain highest
There is no freshwater input into the bay other than
total numbers of fish, calculated from observed
rain, and salinity (avg. 35·4) is slightly higher than on
density and total surface area.
the reef (avg. 34·6). Bay water temperature averaged
In the same clear water marine bay in the
30·1 C ( 0·8), while water temperature on the reef
Caribbean, a selection of seven ‘ nursery species ’ of
averaged 28·4 C ( 0·9). Visibility was high at all
which juvenile individuals had been found in large
sites, and varied between an average of 6·5 m ( 1·8)
numbers in mangroves and seagrass beds was
in the bay and 21·4 m ( 3·1) on the reef as measured
studied in detail (Acanthurus chirurgus, Haemulon
by means of a horizontal Secchi disk. The average
flavolineatum, H. sciurus, Lutjanus apodus, L. griseus,
tidal amplitude in the area is 30 cm (De Haan &
Ocyurus chrysurus, and Scarus iserti). In addition to this
Zaneveld, 1959).
set of herbivorous, zoobenthivorous and piscivorous
The bay has a long (1 km) and narrow ( 70 m)
fish, two congeneric species were selected that
entrance that connects it to the adjacent fringing reef.
were encountered in significant quantities in some
This reef is part of a marine park that stretches up to
seagrass beds near the adjacent fringing reef (Scarus
the southwest tip of the island. The reef system starts
taeniopterus and Acanthurus bahianus). Using daytime
with a shallow reef flat (from 2–7 m depth), typically
underwater visual census as a single method to
covered by gorgonians, at the edge of which the
quantify the abundance of the nine selected species
drop-off is located (at 5–10 m depth). Coral cover on
and estimate their size, heterogeneity of the spatial
the drop-off and reef wall is predominated by the
size-frequency distribution of these fish species in reef
stony coral Montastrea annularis. A detailed descrip-
habitats, mangroves and seagrass meadows was
tion of the reefs in the Netherlands Antilles can be
tested. In this way, association of specific size-classes
found in Bak (1975).
of reef fish with specific habitats or spatially separated
sites provides information from which Post-settlement
Life Cycle Migration (PLCM) patterns can be
Sampling design
derived, taking day-to-day variation over a five month
period into account. Additionally, differences in A total of 35 permanent transects were used in 11 sites
spatial distribution and habitat preference can be on the reef and in the bay (Table 1), covering a total
area of about 4500 m2. Each of the transects was
compared among species.
The questions that will be addressed are: censused 29 times on average, during daytime in May
1. Do size-classes of the selected species display any through to September 1998. In Spanish Water Bay,
preference for mangrove or seagrass habitats in six seagrass sites were selected (Figure 1). At each
terms of densities? seagrass site, three permanent 3 by 50 m belt transects
2. Do habitats differ in the size-structure of the were placed, which were surveyed by snorkelling.
subpopulations that they harbour? Average water depth of these transects was between
3. Which spatial migration patterns can be inferred 0·8 and 2·4 m. Adjacent to four of the seagrass sites
from average densities and sizes by comparison of (numbered 2, 3, 4 and 6) was a mangrove site (Figure
the subpopulations at the various sites? 1). The mangrove stands consist of strips of vegetation
Migration patterns and habitat preference of coral reef fishes 311
Curaçao
N
25°
N
20°
Trade wind
Caribbean Sea
15°
10°
Spanish Water Bay
90° 85° 80° 75° 70° 65° 60°
5m
6
5
m
5m
5m
5 5m
3
10 m
10 m
4
10 m
5m
2
Mangroves
0 250 500 1
Seagrass beds
m
Coral reef
7
Spanish Water Bay
F 1. Location of the study sites in Spanish Water Bay. At sites 1–6 the seagrass beds were censused, while at site 2, 3,
4, and 6 mangroves were also surveyed. Site 7 was the reef site.
hanging over from fossil reef ledges, hence providing (Table 1). At the reef site (numbered 7, Figure 1),
structural complexity from prop roots or branches in three permanent 3 by 50 m belt transects were placed
the water column beneath the mangrove canopy. The at three depths (5, 10 and 15 m) parallel to the
eight mangrove transects were narrow underwater coastline, using nylon twine. The three 5 m deep
habitats, and were censused by snorkelers. In the transects were placed where the sandy reef flat ends at
mangroves, transect width was between 1·1 and 2·1 m the start of the drop-off, the three 10 m deep transects
312 E. Cocheret de la Moriniere et al.
`
T 1. Surface area of the 35 transects in all sites and transects was censused 29 times on average). Data
habitats. Site numbers correspond to Figure 1. On the coral were logtransformed and analysed in a nested
reef, 3 transects were used at each depth, as indicated
ANOVA (GLM, SPSS 8.0) for unequal sample sizes,
between brackets
where sites were nested in habitats and individual
surveys of the transects were treated as replicates
Width Length Area
(m2) within sites. Multiple comparisons of means within
Site Habitat Transect (m) (m)
habitats (among sites) and among habitats were
analysed using a Tukey HSD Spjotvoll/Stoline test
1–6 Seagrass 1 3·0 50·0 150·0
(Sokal & Rohlf, 1995). Observer effect was tested by a
2 3·0 50·0 150·0
3 3·0 50·0 150·0 one-way ANOVA on each of the 35 transects, with
observer identity as an independent variable and mean
2 Mangrove 1 1·3 37·5 48·8
2 1·3 40·0 52·0 size or total density in the surveys as a dependent
3 1 1·1 73·0 80·3 variable. None of the 35 ANOVAs on observer effect
4 1 1·4 43·0 60·2
produced significant differences (P<0·05) in variance
2 1·3 10·0 13·0
among size or density estimation among observers.
6 1 2·1 38·0 79·8
Since there is a variety of prevailing habitats in
2 1·4 37·0 51·8
3 1·5 31·0 46·5 Spanish Water Bay (mangroves, seagrass beds, algal
beds, channel, fossil reef terrace notches, boulders,
7 Reef 1 (3 at 5 m depth) 3·0 50·0 150·0
2 (3 at 10 m depth) 3·0 50·0 150·0 coral reef; see Nagelkerken et al., 2000a), and the
3 (3 at 15 m depth) 3·0 50·0 150·0 selected species use these shallow habitats as daily
resting sites to which they return every day after
nocturnal migrations to deeper feeding or sleeping
grounds (Nagelkerken et al., 2000b), their daytime
were located on the drop-off whereas the three 15 m density distribution can be viewed as a matter of
deep transects were situated on the reef slope. At each choice. Therefore, the density of a size-class of a fish
depth, the three transects were placed 50 m apart species in mangroves relative to its density in seagrass
from each other. The depth range was based on a beds is viewed as a level of habitat preference. The
pilot-study that showed that the selected species level of preference for either mangroves or seagrass
reached highest densities at depths less than 15 m. beds was tested based on densities of the size-classes
Reef sites were censused by Scuba diving. of each fish species occurring in mangroves and
During visual surveys, individuals of the selected seagrass beds at site numbers 2, 3, 4 and 6 (Figure 1).
species were counted and their sizes estimated in Only these sites were used for analysis of habitat
size-classes of 2·5 cm. Underwater size estimators preference because both seagrass and mangrove
were trained with objects of known size. The three habitats were surveyed at those sites. For each size-
observers censused all transects using an alternating class of each species, the average density in mangroves
system so that any bias in size-estimation is equally at a site was divided by the sum density of that
represented in every transect. The observer effect was size-class in mangroves and seagrass beds at that site.
tested using ANOVA (see ‘ Statistical analysis ’ for These mangrove-to-seagrass preference levels of the
further explanation). All juvenile fishes observed in size-classes of the species at the four sites were then
this study were larger than 1 cm at settlement. Juv- clustered using City-block (Manhattan) distances
enile scarids smaller than 5 cm (TL) could not be (Statistica for Windows 4·5). In the Manhattan dis-
identified in the field. Scarids of sizes smaller than tances measure, the effect of single large differences
5 cm were left out of the data sets of Scarus iserti and (outliers) is dampened (in the Euclidean distance
S. taeniopterus. Juveniles and sub-adults of these two measure, differences are squared).
scarids that were larger than 5 cm (TL) could be
distinguished by the characteristics shown in Humann
Results
(1996). All other species could be identified at all
sizes.
Habitat preference
Mean densities (100 m 2) of most species are signifi-
Statistical analysis
cantly lower on the coral reef than in seagrass or
For each species, mean size (cm) and total density mangrove habitats, with the exception of Ocyurus
(N 100 m 2) of the observed individuals was calcu- chrysurus (coral reef densities similar to densities in
lated at each survey of a transect (each of the 35 seagrass beds and lowest densities in mangroves) and
Migration patterns and habitat preference of coral reef fishes 313
T 2. Average size (cm) and density (N 100 2) per species in each habitat, and their standard errors between brackets.
Among sizes, significant (P<0·05) differences are indicated with a, b, and c for each species. Among densities, significantly
different means are marked d, e, or f for each species. Different letters (a–f) mean that averages are significantly different
Size Densities
Mangrove Seagrass Reef Mangrove Seagrass Reef
4·4a (0·4) 12·9b (0·6) 4·3d (0·5) 4·9d (1·6)
Acanthurus bahianus — 0·0
13·0a (0·3) 11·0b (0·3) 17·0c (0·5) 3·3d (0·3) 0·9e (0·2) 2·8f (0·9)
Acanthurus chirurgus
8·8a (0·2) 7·7b (0·1) 15·1c (0·2) 99·8d (12·2) 32·1e (1·7) 3·2f (0·4)
Haemulon flavolineatum
12·3a (0·2) 11·5b (0·1) 21·5c (0·3) 18·2d (1·2) 5·7e (0·3) 0·3f (0·0)
Haemulon sciurus
12·3a (0·2) 11·3a (0·5) 18·5b (0·5) 24·7d (1·4) 0·3e (0·1) 1·0f (0·1)
Lutjanus apodus
14·2a (0·3) 12·6b (0·3) 16·6a (1·1) 8·1d (0·8) 0·5e (0·1) 0·0f (0·0)
Lutjanus griseus
9·3a (0·4) 9·8a (0·2) 17·1b (0·3) 2·1d (0·4) 3·6e (0·4) 4·1e (0·5)
Ocyurus chrysurus
8·1a (0·3) 7·8a (0·1) 11·9b (0·3) 11·5d (2·0) 13·3d (1·3) 2·9e (0·2)
Scarus iserti
6·3a (0·0) 6·4a (0·1) 16·4b (0·3) 6·3d (1·3) 1·7e (0·3) 3·6f (0·2)
Scarus taeniopterus
Acanthurus bahianus (not observed in mangroves, and ever-larger sizes found in clusters with an increasing
seagrass densities not significantly different from reef mangrove-to-seagrass density ratio. The latter two
densities) (Table 2). Overall densities of Scarus iserti, species suggest that their smallest juveniles are most
Ocyurus chrysurus, and Acanthurus bahianus are higher commonly found in seagrass beds, while mangrove
in seagrass beds than in mangroves, while the reverse preference increases with size of the juveniles. Only
is true for the remaining species (Table 2). one size-class (5·0–7·5 cm) of Scarus taeniopterus was
In order to determine the level of habitat observed in the mouth of the bay, of which indi-
preference, mangrove-to-seagrass density ratios (see viduals were observed in mangroves and seagrass beds
Statistical analysis) were determined for each size-class (cluster B).
at four sites where both mangroves and seagrass beds
were surveyed. Cluster analysis of these mangrove-to-
Size-distribution over habitats
seagrass density ratios at the four sites (numbered 2,
3, 4, and 6 in Figure 1) yielded three distinct groups of In Figure 3, relative densities of the selected species in
size-classes of fishes at linkage distance 2·0 (Figure 2). the three habitats are depicted for each size-class.
Cluster A had an average mangrove-to-seagrass All selected species were ‘ nursery species ’, in the
density ratio of 18% (range 0–46%); in cluster B, that sense that high densities of juveniles were found in
ratio is 60% (34–75%), and in cluster C it is 93% mangroves or seagrass beds, while most adults were
(64–100%). All size-classes of Lutjanus apodus and observed on the reef. All three habitats differed
Haemulon sciurus belonged to cluster C, reflecting a significantly in the average sizes of individuals of
strong preference for mangroves at all size-classes. Haemulon flavolineatum, H. sciurus, and Acanthurus
Lutjanus griseus also seems to prefer mangroves over chirurgus that they harboured (Table 2). Ocyurus
seagrass beds, since all size-classes were members of chrysurus, Lutjanus apodus, Scarus iserti, and S.
cluster C, and one size-class (7·5–10 cm) was in taeniopterus showed no difference in average sizes
cluster B. Acanthurus bahianus was only observed in between mangroves and seagrass beds, but individuals
seagrass habitats, and therefore all size-classes were on the reef were significantly larger. Acanthurus
members of cluster A. Acanthurus chirurgus distributed bahianus was never found in mangroves, and the
itself over mangroves and seagrass beds (most size- average size of the individuals of this species observed
classes are members of cluster B), with a few individ- in seagrass beds was smaller than on the reef. Average
uals of the largest size-class found only in mangroves size of Lutjanus griseus was significantly smaller in
(and therefore part of cluster C). Ocyurus chrysurus seagrass beds than in mangroves and on the coral reef,
was observed in both seagrass beds and mangroves while the latter two habitats showed average sizes that
(size-classes are part of clusters A and B). Scarus iserti were similar to each other. Mean size of all nine
occurred mostly in seagrass beds (cluster A), but species on the reef (Table 2) was always smaller than
mangrove preference seems to increase with size or corresponded to the approximate mean total
(largest sizes in cluster B). Haemulon flavolineatum is lengths at which these species become sexually mature
represented by size-classes in all three clusters, with (see Figure 3).
314 E. Cocheret de la Moriniere et al.
`
Linkage distance
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
O. chrys 17.5–20
O. chrys 2.5–5
O. chrys 15–17.5
A. bah 12.5–15
A. bah 10–12.5
A. bah 7.5–10
A. bah 5–7.5
A. bah 2.5–5 A
A. bah 0–2.5
S. iser 7.5–10
S. iser 5–7.5
O. chrys 5–7.5
O. chrys 7.5–10
S. iser 10–12.5
H. flav 5–7.5
H. flav 2.5–5
H. flav 0–2.5
O. chrys 20–22.5
O. chrys 0–2.5
S. taen 5–7.5
O. chrys 12.5–15
B
O. chrys 10–12.5
H. flav 10–12.5
H. flav 7.5–10
A. chir 10–12.5
L. gris 7.5–10
S. iser 12.5–15
A. chir 15–17.5
A. chir 7.5–10
A. chir 12.5–15
S. iser 15–17.5
A. chir 20–22.5
A. chir 5–7.5
A. chir 2.5–5
A. chir 17.5–20
A. chir 0–2.5
L. apo 5–7.5
L. apo 2.5–5
L. apo 7.5–10
L. gris 15–17.5
L. apo 12.5–15
L. apo 10–12.5
L. apo 15–17.5
L. gris 17.5–20
L. apo 17.5–20
H. flav 15–17.5
L. apo 25–27.5
L. apo 22.5–25
L. apo 20–22.5
L. apo 0–2.5
L. gris 20–22.5
A. chir 22.5–25
H. sci 17.5–20
L. gris 22.5–25
H. sci 20–22.5
L. gris 12.5–15
L. gris 10–12.5 C
L. gris 5–7.5
H. flav 12.5–15
H. sci 15–17.5
H. sci 12.5–15
H. sci 10–12.5
H. sci 7.5–10
H. sci 5–7.5
H. sci 2.5–5
H. sci 0–2.5
F 2. Complete linkage of relative (mangrove-to-seagrass) densities of the size-classes of the study species in four sites
(numbered 2, 3, 4 and 6; see Figure 1), using City-block (Manhattan) distances. Species are indicated by the following codes:
A. bah=Acanthurus bahianus, A. chir=Acanthurus chirurgus, H. flav=Haemulon flavolineatum, H. sci=Haemulon sciurus,
L. apo=Lutjanus apodus, L. gris=Lutjanus griseus, O. chrys=Ocyurus chrysurus, S. iser=Scarus iserti, S. taen=Scarus taeni-
opterus. Size-classes (cm) are indicated by the numbers behind the species codes.
Migration patterns and habitat preference of coral reef fishes 315
Spatial migration patterns of a fish species over mangroves and seagrass beds at
daytime is considered here as a matter of choice.
The size-frequency distribution of Haemulon
Cluster-analysis of the mangrove-to-seagrass density
flavolineatum, H. sciurus, Acanthurus chirurgus,
ratios of each size-class, at four sites that harboured
Lutjanus apodus, Scarus iserti, and Ocyurus chrysurus in
both habitats showed different levels of habitat pref-
mangroves and seagrass beds and on the reef suggests
erence. Lutjanus apodus and Haemulon sciurus showed
a size-range for each species over which the juveniles
strong preference for mangroves over seagrass beds at
start migrating to the coral reef (Figure 3). In the case
all size-classes. L. griseus was also strongly associated
of Scarus taeniopterus this migration from nursery
with mangroves at all size-classes, and moderately by
habitat to coral reef appears to take place rather
one size-class. Acanthurus bahianus was not observed
abruptly, while individuals of Acanthurus bahianus
in mangroves, reflecting strong preference for seagrass
may migrate to reef habitats at all sizes.
beds. Ocyurus chrysurus and Acanthurus chirurgus
When size-frequencies were compared among bay
utilized both habitats. Scarus iserti and Haemulon
sites within each habitat (ANOVA), heterogeneous
flavolineatum also used both habitats, but there was a
distribution patterns over sites emerged for some
trend of increased preference for mangroves with
species (Table 3). For H. sciurus, the extremely low
increasing fish size, while the smallest juveniles of
(0·09 100 m 2) densities of small individuals at site 2
these species were highly associated with seagrass
are responsible for significant differences among
beds. Though seemingly marginal habitats, strips of
seagrass sites.
mangroves of no more than 1 by 40 m at times
Haemulon flavolineatum, Scarus iserti, and Ocyurus
may contain hundreds of individuals in resting
chrysurus displayed a size-frequency distribution in
schools. The preference of Scarus taeniopterus for
which high densities of small individuals were found
mangroves may be exaggerated (mangrove-to-seagrass
in mangroves and seagrass beds in the mouth of the
density ratio was about 60% : 40%), since juveniles
bay (at site numbers 1 and 2, Figure 1), medium-sized
smaller than 5 cm were excluded from the data set.
fishes deeper in the bay, and large fishes on the reef
Unidentifiable scarid juveniles of this size were mostly
[Figure 4(a, b and c)]. This indicates a Post-
found in the mouth of the bay (average density
settlement Life Cycle Migration (PLCM) pattern with
38·2 100 m 2 in the mouth of the bay as opposed to
two changes of direction.
0·8 100 m 2 in transects deeper in the bay) in sea-
Both Acanthurus bahianus and Scarus taeniopterus
grass beds. Nagelkerken et al. (2000a, c) have found
were only encountered at sites 1 and 2, which are
similar overall density distributions of these species in
located in the mouth of the bay (see Figure 1), and on
mangrove habitats and seagrass habitats in Curacao ¸
the reef. They were not observed at sites located
and Bonaire. The level of preference of these fish
deeper into the bay. In both cases, high densities of
species for mangroves or seagrass beds in the situation
small juveniles were detected in the mouth of the bay
where both habitats occur, however, is no indication
[Figure 4(d and e)], while statistically lower densities
of the level of dependence on these habitats. From
and larger individuals of these species occurred on the
comparisons among bays with and without mangroves
reef (Table 2). Their distribution indicates a PLCM
or seagrass beds (Nagelkerken et al., 2001), it is
pattern that is restricted to seagrass and mangrove
known that species that showed strong preference for
sites in the close vicinity of the reef, and does not
mangroves in the present study (Lutjanus apodus, L.
include temporary residence deeper in the bay.
griseus, Haemulon sciurus) depend largely on the pres-
ence or absence of seagrass beds. Given the choice,
such species apparently prefer mangroves as daytime
Discussion
resting sites for shelter, while their dependence on
seagrass beds is best explained by the larger abun-
Habitat preference
dance of food in seagrass habitats in which they forage
From previous studies in the same bay (Nagelkerken at night.
et al., 2000a, b), it is known that fish have a number of
occurring habitats to choose from in Spanish Water
Size-distribution and spatial migration patterns
Bay (mangroves, seagrass beds, algal beds, channel,
fossil reef terrace notches, boulders). These shallow All selected species proved to be ‘ nursery species ’ in
habitats are used as daily resting sites, to which the the sense that juveniles were much more abundant in
fishes return every day after nocturnal migrations to mangroves or seagrass beds than on the reef, as
deeper feeding or sleeping grounds (Nagelkerken et expected from our previous study (Nagelkerken et al.,
al., 2000b). Therefore, the relative density distribution 2000a). Of nine species, six (Haemulon flavolineatum,
Average density (%) Average density (%)
Average density (%) Average density (%) Average density (%)
0. 0.
5. 0. 0.
0
0
0–
0
20
40
60
80
100
0– 0–
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
–7
2. 2. –2. 2. 2.
55
.5 5 5
2.
2. 5 2.
5–
5– 5. –5.
7. 5–
00 5.
5. 5–
0 5.
5.
7. –7 0
5.
10
5. 0 0– 5.
.0
0– 10 –1 5 7. 0–
7.
.0
10
7.
5 5– 5 7.
.0 5
7. 12 0–1 .0 7.
Seagrass
.2 10 10
–1
5–
Seagrass
Mangrove
5–
.0 .0
2.
10 15 5–1 .5 –
5 10
Seagrass
.0
10 12 .0 5. .0
Seagrass
12 12 10
.0 .5 17 –1 0
Mangrove
Mangrove
.5 .5 .0
.7 –
–1 –1
`
Seagrass
–1
20 5–2 .5
5. 15 15
12 2.5
0 2.
12 5
.0 .0
.5 .0 0.
15
– .5
22 –2 0
.0
–1
.2 –1
17 17
–1
Mangrove
15 5.0 .5 .5 5.
25 5–2 .5 15 0
7. –
.0 .5
5 .0
316 E. Cocheret de la Moriniere et al.
17
–1 20 20
27 0–2 .0
(e) L. apodus
–1
(i) A. bahianus
.5 .0 .0
(c) A. chirurgus
Size-classes (cm)
17 7.5
Reef
(g) Scarus iserti
–
.5 7. 7.
17
–2 5
.5 30 –3 5 .5
22 22
(a) H. flavolineatum
.0
0.
–2 0 .5 .5 –2
20
–
Reef
32 0–3 .0 Reef
20 0.0
.2 0.
.0 20 0
.0 25 25 .0
–2 35 5–3 .5 .0 .0
–2
.5 – –2
2.
Reef
5
22 2.5 2.
Reef
37 0–3 .0 27 27
22 22 5
.5 .5 .5
.5 .5
.5 7.
–2 –4 5 –3
–2 –2
5.
0 0. 0.
5. 5.
0 0
0 0
Average density (%) Average density (%)
Average density (%) Average density (%)
0. 0.
5. 5.
0– 0–
0– 0–
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
2.
7. 7.
2. 2.5 5
5 5 2.
7.
5– 7. 5–
5–
5–
5. 5.0 5.
0 0
10
10 5.
.0
10
.0 0–
7. –7. 10 .0
5– 5 7.
7.
.0 –1 5– 5
10 10 –1 2.
12
.0 .0 5
Mangrove
Seagrass
10 10
2. .5
5
12 .0 .0
12 –1
Seagrass
–
–1
.5
.5 2.
Seagrass
5.
–1 12 12
15 0
15 –1 5 .5 .5
Mangrove
.0
5.
.0 5. 0 –
15 –1
.0 15 15
17 –1 0
Seagrass
7.
17 .0 .0
–1
.5 7.5 .5 5 –1
7.
20 –2 5 –2
17 17 7
.0 0. .5 .5 .5
20 0.0
–
22 –2 0 –2
Mangrove
.0
.5 2. 20 20
(f) L. griseus
0. –2
Mangrove
(b) H. sciurus
0 .0 .0
20
25 –2 5 –
Size-classes (cm)
.0 22 2.5
Reef
.0 5. .5 22 22
Reef
–2
(h) Ocyurus chrysurus
.5 .5
27 –2 0 –2
(d) Scarus taeniopterus
2.
–
5
.5 7. 22 25 5.0
.5 25 25
.0
30 –3 5 .0 .0
–2
Reef
.0 0. –
–2
5.
0 27 27
32 –3 0 25
Reef
27 7.5
.5 .5
.0 .5
.5 2.5
–3
–3 –2 –3
0.
5. 7. 0.
0
0 5 0
Migration patterns and habitat preference of coral reef fishes 317
T 3. Analysis of variance of sizes among sites within bay assumed that it is a ‘ nursery species ’ as well, and
habitats. Bold print represents P-values smaller than 0·05. that it migrates according to the above mentioned
ANOVA could not be applied to A. bahianus and S.
Long Distance PLCM pattern. This distribution
taeniopterus. Significant P-values indicate that variance of
pattern fits the observed size-distribution by other
sizes of fishes is not homogeneously distributed over sites
authors well (e.g. Parrish, 1989 and references
within that habitat
therein; Rutherford et al., 1989; Appeldoorn et al.,
Mangrove Seagrass 1997), with larger individuals found progressively
off-shore.
Three species, Haemulon flavolineatum, Ocyurus
Acanthurus bahianus — 0·44
Acanthurus chirurgus 0·20 0·16 chrysurus and Scarus iserti, showed similar size-
Haemulon flavolineatum 0·00 0·00 frequency distribution patterns with three spatially
Haemulon sciurus 0·20 0·00
separated groups that were statistically different: small
Lutjanus apodus 0·15 0·16
juveniles in the mouth of the bay, larger individuals
Lutjanus griseus 0·35 0·27
in seagrass beds and mangroves located deeper into
Ocyurus chrysurus 0·10 0·00
Scarus iserti 0·00 0·00 the bay and (sub-) adults on the reef. This spatial
Scarus taeniopterus — 0·00 distribution suggests a Life Cycle Migration that
involves more than one direction of migration, and is
hence termed Stepwise PLCM. Post-larvae of these
species settle in the mouth of the bay, after which they
H. sciurus, Lutjanus apodus, Ocyurus chrysurus, migrate deeper into the bay to grow up to a size large
Acanthurus chirurgus, and Scarus iserti) showed enough to migrate to and dwell on the reef. Most
spatial distributions in which smallest individuals migrants into coastal regions come from the open sea
were only found in bay habitats. The largest indi- (Blaber, 1997). These are in-out migrations: juveniles
viduals of Haemulon flavolineatum, Lutjanus apodus, or adults or both migrate into an estuarine or coastal
Ocyurus chrysurus, and Scarus iserti were found only on area for a certain period, after which they return to the
the adjacent reef, while adults of Haemulon sciurus, open sea or coral reef. De Sylva (1963) describes a
and Acanthurus chirurgus were found both in reef distribution pattern of Sphyraena barracuda that is
habitats as well as in bay habitats (mangrove or coherent to the Stepwise PLCM patterns that were
seagrass). Average size of these six species was largest found in our study. Post-larvae and juveniles of this
on the reef. The size at which these species become piscivore move from coastal shallows to reed beds or
sexually mature (Robertson & Warner, 1978; Munro, mangroves, followed by a migration to open sea.
1983; Munro pers. comm.) always corresponded to or Juveniles of Scarus taeniopterus and Acanthurus
was larger than the average size at which they were bahianus were only found in the mouth of the bay and
found on the coral reef. The results suggest a Post- on the reef. Small juveniles of Acanthurus bahianus
settlement Life Cycle Migration (PLCM) pattern over were only observed in seagrass beds in the mouth of
a considerable distance, in which juveniles settle the bay and in the reef flats, while adults occurred
and grow up in alternative habitats such as seagrass almost exclusively on the reef. These two species
beds and mangroves, after which the sub-adults display a type of Short Distance PLCM in which
migrate to reef habitats where they become sexually larvae partly settle in the mouth of the bay and partly
mature. That pattern is named Long Distance in the reef flats, to reach a size at which they migrate to
PLCM. Average size of Lutjanus griseus was smallest in deeper reef habitats.
seagrass beds, while average sizes in mangroves and Interestingly, some congeneric species appeared to
on reefs were similar. The size at which Lutjanus display different directions of migration at similar
griseus becomes sexually mature is about 25 cm sizes. Scarids of similar sizes are found mixed in the
(Starck, 1971; Claro, 1983), and individuals of this mouth of the bay, but one species then migrates to
size have been observed in mangroves as well as on reef habitats (Scarus taeniopterus), and the other
the reef. Since small juveniles of Lutjanus griseus migrates deeper into the bay and only dwells on the
were only observed in bay habitats, and spawning reef at larger sizes (Scarus iserti). Leaving unidenti-
occurs on the reef or shelf edge (Claro, 1983), it is fiable scarids smaller than 5 cm out of the data sets
F 3. Average relative density of individuals of the selected species in three habitats, calculated per size-class. The
arrows indicate approximate length at sexual maturity (maturity data from Munro, 1983; Starck, 1971; Claro, 1983).
318 E. Cocheret de la Moriniere et al.
`
18 17
(a) (b)
16
15
14
Mean size (cm)
13
12
d
cd c
10 11
ce e c
b
b b
a b
8
b 9
6 a
b
a b
a a
7
4
a
2 5
M2 M3 M4 M6 S1 S2 S3 S4 S5 S6 Reef M2 M4 M6 S1 S2 S3 S4 S5 S6 Reef
Sites Sites
24 24
(c) (d)
20 20
Mean size (cm)
16 16
bc
b cd
d
12 12
a
b
a
b
8 8
a a b
ab
4 4
0 0
M2 M3 M4 M6 S1 S2 S3 S4 S5 S6 Reef M2 S1 S2 Reef
Sites Sites
20
(e)
18
16
Mean size (cm)
14
12 ± Std. Dev.
10 Mean
8
a
a
6
4
2
S1 S2 Reef
Sites
F 4. Mean and standard deviation of size-distribution of Haemulon flavolineatum (a), Scarus iserti (b), Ocyurus chrysurus
(c) Scarus taeniopterus (d) and Acanthurus bahianus (e) at each site. Within each habitat (indicated with ‘ S ’, ‘ M ’, or ‘ Reef ’,
for seagrass, mangrove and reef habitats, respectively), statistically significant differences among sites (Tukey HSD) are
presented by letters only when the species was found in that habitat at more than one site. Different letters mean that the
averages of the sites are statistically different.
did not affect conclusions regarding the spatial distribution, abundance and early post-settlement
migration patterns of these species, since this size- persistence of settlers (Risk, 1998), while post-
class was almost exclusively found in the mouth of settlement habitat selection is important in creating
the bay. Apparently, both scarids settle in seagrass spatial patterns of recruitment (Sponaugle & Cowen,
beds and mangroves located in the mouth of the bay, 1996). This means that competitive congeneric
after which each migrates in an opposite direction. species can alleviate competition on the reef by
The same difference is observed when comparing the temporary spatial separation.
migration patterns of Acanthurus chirurgus and A. Possible explanations for different spatial size-
bahianus. Of Acanthurus bahianus it is known that frequency distributions of post-settlement fishes
behavioural interactions are size-related and can affect involve variability in mortality rates, growth,
Migration patterns and habitat preference of coral reef fishes 319
F 5. Three types of migration among nurseries and the coral reef. Route 1 depicts the Short Distance PLCM, route 2
Stepwise PLCM, and route 3 Long Distance PLCM (explanation see text).
settlement patterns and migration patterns. Since the Conclusions
abundance of predators is much lower in the bay
than on the reef as is generally the case (Shulman, Taking day-to-day variations in fish density and size-
1985; Parrish, 1989), differences in mortality rates frequency into account over a five-month period,
may explain the high abundance of juveniles in bay spatial patterns emerge for the selected fish species. Of
habitats and the reduced numbers on the coral reef. some species, all size-classes that occurred in bay
In fact, reduced mortality among juveniles in nursery habitats appeared to prefer mangroves as daytime
habitats is often ascribed to reduced predator abun- resting sites, while others were only found in seagrass
dance or efficiency (e.g. Heck & Crowder, 1991; beds. Other species utilized and preferred mangroves
Robertson & Blaber, 1992). This, however, cannot and seagrass beds at different sizes, and preference for
explain the lower number of (sub-)adults in man- mangroves of some species increased with increasing
grove and seagrass habitats or the low numbers of size-class.
the smallest juveniles in the habitats that are located The size-frequency distribution patterns of
deep in the bay. Also, abundance and availability of Haemulon flavolineatum, H. sciurus, Acanthurus
food items (such as benthic and planktonic inverte- chirurgus, Lutjanus apodus, Scarus iserti, and Ocyurus
brates, epifauna and epiphytes) is much higher in the chrysurus in mangroves and seagrass beds and on
bay habitats of Spanish Water Bay than on the the reef suggest a size-range for each species over
nearby coral reef (Cocheret de la Moriniere et al.,
` which the juveniles start migrating to the coral reef
unpublished), which could not result in lower (Figure 3). In the case of Scarus taeniopterus this
growth rates of fishes in the bay. Variability in migration from nursery habitat to coral reef appears to
growth rate is therefore another unlikely explanatory take place rather abruptly, while individuals of Acan-
factor for the fact that the largest individuals of thurus bahianus may migrate to reef habitats at all
nursery species are usually found on the coral reef. sizes.
The spawning seasons of the selected species are Haemulon sciurus, Lutjanus apodus, L. griseus, and
largely during the study period (Munro et al., 1973), Acanthurus chirurgus display Long Distance PLCM;
and regular settlement (no major peaks) of post- Haemulon flavolineatum, Ocyurus chrysurus and
larvae was observed for most of the species during Scarus iserti use Stepwise PLCM; Scarus taeniopterus
the study. Considering all these processes, migration and Acanthurus bahianus are retained within a small
from nursery ground to coral reef habitat seems a distance from the reef (Short Distance PLCM).
logical explanation for the spatial distribution of These different migration patterns are depicted in
size-classes of these fish species, and migratory pat- Figure 5. The fact that some species carry out specific
terns can be inferred. The stability of such patterns directional migrations and congeners may migrate
and validity of actual migrations must be tested in to different areas raises questions concerning the
further studies. mechanisms that trigger these migrational options,
320 E. Cocheret de la Moriniere et al.
`
Heck, K. L. & Crowder, L. B. 1991 Habitat structure and predator-
and their ecological or evolutionary meaning. This
prey interactions in vegetated aquatic ecosystems. In Habitat
also encourages scientists to view the ‘ nursery ques- Structure: The Physical Arrangements of Objects in Space (Bell, S. S.,
tion ’ not only from the point of view of benevolence McCoy, E. D. & Mushinsky, E. R., eds). Chapman and Hall,
London, pp. 281–299.
of nursery areas, but also to elucidate why a species
Humann, P. 1996 Reef Fish Identification, 2nd edition. Florida
shows a particular migration pattern or why it does Caribbean Bahamas. New World Publications Inc., Jacksonville,
not. USA, 396 pp.
Kuenen, M. M. C. E. & Debrot, A. O. 1995 A quantitative study of
the seagrass and algal meadows of the Spaanse Water, Curacao, ¸
The Netherlands Antilles. Aquatic Botany 51, 291–310.
Acknowledgements Laegdsgaard, P. & Johnson, C. 2001 Why do juvenile fish utilise
mangrove habitats? Journal of Experimental Marine Biology and
We would like to thank all personnel of the Ecology 257, 229–253.
Carmabi Foundation, where the research was carried Munro, J. L., Gaut, V. C., Thompson, R. & Reeson, P. H. 1973
The spawning seasons of Caribbean reef fishes. Journal of Fish
out. The research was funded by the Netherlands Biology 5, 69–84.
Foundation for the Advancement of Tropical Munro, J. L. (ed.) 1983 Caribbean Coral Reef Fishery Resources, 2nd
Research (WOTRO). The Stichting Nijmeegs edition. ICLARM, Philippines, 276 pp.
Nagelkerken, I., Dorenbosch, M., Verberk, W. C. E. P., Cocheret
Universiteitsfonds (SNUF) funded B. J. A. Pollux.
de la Moriniere, E. & van der Velde, G. 2000a Importance of
`
Furthermore, we thank E. Kardinaal for supplying us shallow-water biotopes of a Caribbean bay for juvenile coral reef
with a map of the Spanish Water Bay and the Winkel fishes: patterns in biotope association, community structure
and spatial distribution. Marine Ecology Progress Series 202, 175–
family for the use of their pier. Finally, we would like
192.
to thank Blu Forman for her help with the graph on Nagelkerken, I., Dorenbosch, M., Verberk, W. C. E. P., Cocheret
migration patterns, and P. H. Nienhuis for useful de la Moriniere, E. & van der Velde, G. 2000b Day-night shifts
`
comments on the manuscript. of fishes between shallow-water biotopes of a Caribbean bay,
with emphasis on the nocturnal feeding of Haemulidae and
Lutjanidae. Marine Ecology Progress Series 194, 55–64.
Nagelkerken, I., van der Velde, G., Gorissen, M. W., Meijer, G. J.,
References van ’t Hoff, T. & den Hartog, C. 2000c Importance of man-
groves, seagrass beds and the shallow coral reef as a nursery for
Appeldoorn, R. S., Recksiek, C. W., Hill, R. L., Pagan, F. E. & important coral reef fishes, using a visual census technique.
Dennis, G. D. 1997 Marine protected areas and reef fish move- Estuarine, Coastal and Shelf Science 51, 31–44.
ments: the role of habitat in controlling ontogenetic migration.
Nagelkerken, I., Kleijnen, S., Klop, T., van den Brand, R. A. C. J.,
Proceedings of the 8th International Coral Reef Symposium 2,
Cocheret de la Moriniere, E. & van der Velde, G. 2001 Depen-
`
1917–1922.
dence of Caribbean reef fishes on mangroves and seagrass beds as
Baelde, P. 1990 Differences in the structures of fish assemblages in
nursery habitats: a comparison of fish faunas between bays with
Thalassia testudinum beds in Guadeloupe, French West Indies,
and without mangroves and seagrass beds. Marine Ecology
and their ecological significance. Marine Biology 105, 163–173.
Progress Series 214, 225–235.
Bak, R. P. M. 1975 Ecological aspects of the distribution of reef
Odum, W. E. & Heald, E. J. 1972 Trophic analyses of an estuarine
corals in the Netherlands Antilles. Bijdragen tot de Dierkunde 45,
mangrove community. Bulletin of Marine Science 22, 671–738.
181–190.
Ogden, J. C. & Ehrlich, P. R. 1977 The behaviour of heterotypic
Birkeland, C. 1985 Ecological interactions between tropical coastal
resting schools of juvenile grunts (Pomadasyidae). Marine Biology
ecosystems. UNEP Regional Seas Reports and Studies 73, 1–26.
42, 273–280.
Blaber, S. J. M., Brewer, D. T. & Salini, J. P. 1989 Species
Ogden, J. C. & Gladfelter, E. H. 1983 Coral reefs, seagrass beds,
composition and biomasses of fishes in different habitats of a
and mangroves: their interactions in the coastal zones of the
tropical northern Australian estuary: their occurrence in the
Caribbean. UNESCO Reports in Marine Science 23, 133 pp.
adjoining sea and estuarine dependence. Estuarine, Coastal and
Ogden, J. C. & Zieman, J. C. 1977 Ecological aspects of coral
Shelf Science 29, 509–531.
reef-seagrass bed contacts in the Caribbean. Proceedings of the 3rd
Blaber, S. J. M. (ed.) 1997 Fish and Fisheries of Tropical Estuaries,
International Coral Reef Symposium 1, 377–382.
Volume 22. Chapman & Hall, London, 353 pp.
Parrish, J. D. 1989 Fish communities of interacting shallow-water
Blaber, S. J. M. & Blaber, T. G. 1980 Factors affecting the
habitats in tropical oceanic regions. Marine Ecology Progress Series
distribution of juvenile estuarine and inshore fish. Journal of Fish
58, 143–160.
Biology 17, 143–162.
Pollard, D. A. 1984 A review of ecological studies on seagrass-fish
Carr, W. E. S. & Adams, C. A. 1973 Food habits of juvenile marine
communities, with particular reference to recent studies in
fishes occupying seagrass beds in the estuarine zone near Crystal
Australia. Aquatic Botany 18, 3–42.
River, Florida. Transactions of the American Fisheries Society 102,
Risk, A. 1998 The effects of interactions with reef residents on
511–540.
the settlement and subsequent persistence of ocean surgeon-
Claro, R. 1983 Ecologia y cicle de vida del caballerote,
fish, Acanthurus bahianus. Environmental Biology of Fishes 51,
Lutjanus griseus (Linnaeus), en la platforma cubana. I. Identidad,
377–389.
distribucion y habitat, nutricion y reproduccion. Reportajes y
Investigaciones del Instituto por Oceanologia de la Academia de Robertson, A. I. & Duke, N. C. 1987 Mangroves as nursery sites:
Sciencia de Cuba 7, 1–30. comparisons of the abundance and species composition of fish
and crustaceans in mangroves and other nearshore habitats in
De Haan, D. & Zaneveld, J. S. 1959 Some notes on tides in
tropical Australia. Marine Biology 96, 193–205.
Annabaai harbour, Curacao, Netherlands Antilles. Bulletin of
¸
Marine Science of the Gulf and Caribbean 9, 224–236. Robertson, A. I. & Blaber, S. J. M. 1992 Plankton, epibenthos and
De Sylva, D. P. 1963 Systematics and life-history of the great fish communities. In Tropical Mangrove Ecosystems (Robertson,
barracuda Sphyraena barracuda. Studies in Tropical Oceanography A. I. & Alongi, D. M., eds). Coastal and Estuarine Studies
1, 179 pp. No. 41, 173–224.
Migration patterns and habitat preference of coral reef fishes 321
Robertson, D. R. & Warner, R. R. 1978 Sexual patterns in the Sponaugle, S. & Cowen, R. K. 1996 Larval supply and patterns
labroid fishes of the Western Caribbean, II: The parrotfishes of recruitment for two Caribbean reef fishes, Stegastus partitus
(Scaridae). In Smithsonian Contributions to Zoology (Robertson, and Acanthurus bahianus. Marine and Freshwater Research 47, 433–
D. R. & Warner, R. R., eds). Smithsonian Institute, Washington, 447.
DC, No. 255, 635 pp. Starck, W. A. & Schroeder, R. E. 1971 Investigations on the gray
Rooker, J. R. & Dennis, G. D. 1991 Diel, lunar and seasonal snapper Lutjanus griseus. Studies in Tropical Oceanography 10,
changes in a mangrove fish assemblage off southwestern Puerto 224 pp.
Rico. Bulletin of Marine Science 49, 684–698. Thayer, G. W., Colby, D. R. & Hettler, W. F. Jr 1987 Utilization of
Rutherford, E. S., Schmidt, T. W. & Tilman, J. T. 1989 Early life the red mangrove prop root habitat by fishes in south Florida.
history of spotted sea trout (Cynoscion nebulosus) and gray snapper Marine Ecology Progress Series 35, 25–38.
(Lutjanus griseus) in Florida Bay, Everglades National Park, Weinstein, M. P. & Heck, K. L. Jr 1979 Ichthyofauna of seagrass
Florida. Bulletin of Marine Science 44, 49–64. meadows along the Caribbean coast of Panama and in the gulf of
´
Sedberry, G. R. & Carter, J. 1993 The fish community of a shal- Mexico: composition, structure and community ecology. Marine
low tropical lagoon in Belize, Central America. Estuaries 16, Biology 50, 97–107.
198–215. Yanez-Arancibia, A., Lara-Dominguez, A. L., Rojas-Galaviz, J. L.,
´˜
Shulman, M. J. 1985 Recruitment of coral reef fishes: efects of Sanchez-Gil, P., Day, J. W. & Madden, C. J. 1988 Seasonal
´
distribution of predators and shelter. Ecology 66, 1056–1066. biomass and diversity of estuarine fishes coupled with tropical
Sokal, R. R. & Rohlf, F. J. 1995 Biometry, 3rd edition. WH habitat heterogeneity (southern Gulf of Mexico). Journal of Fish
Freeman and Company, New York, 887 pp. Biology 33 (Suppl A), 191–200.
doi:10.1006/ecss.2001.0907, available online at http://www.idealibrary.com on
Post-settlement Life Cycle Migration Patterns and
Habitat Preference of Coral Reef Fish that use
Seagrass and Mangrove Habitats as Nurseries
E. Cocheret de la Morinierea, B. J. A. Polluxa, I. Nagelkerkena and
`
G. van der Veldea,b
a
Department of Animal Ecology and Ecophysiology, University of Nijmegen, Toernooiveld 1, 6525 ED Nijmegen,
The Netherlands
Received 14 May 2001 and accepted in revised form 8 October 2001
Mangroves and seagrass beds have received considerable attention as nurseries for reef fish, but comparisons have often
been made with different methodologies. Thus, relative importance of different habitats to specific size-classes of reef fish
species remains unclear. In this study, 35 transects in 11 sites of mangroves, seagrass beds and coral reef were surveyed
daily, in and in front of a marine bay on the island of Curacao (Netherlands Antilles). The density and size-frequency of
¸
nine reef fish species (including herbivores, zoobenthivores and piscivores) was determined during a five-month period
using a single methodology, viz. underwater visual census. All species were ‘ nursery species ’ in terms of their high
densities of juveniles in mangroves or seagrass beds. Relative density distribution of the size-classes of the selected species
over mangroves and seagrass beds suggested high levels of preference for either mangroves or seagrass beds of some
species, while other species used both habitats as a nursery. Spatial size distribution of the nine species suggested three
possible models for Post-settlement Life Cycle Migrations (PLCM). Haemulon sciurus, Lutjanus griseus, L. apodus, and
Acanthurus chirurgus appear to settle and grow up in bay habitats such as mangroves and seagrass beds, and in a later stage
migrate to the coral reef (Long Distance PLCM). Juveniles of Acanthurus bahianus and Scarus taeniopterus were found only
in bay habitats at close proximity to the coral reef or on the reef itself, and their migration pattern concerns a limited
spatial scale (Short Distance PLCM). Some congeneric species carry out either Long Distance PLCM or Short Distance
PLCM, thereby temporarily alleviating competition in reef habitats. Haemulon flavolineatum, Ocyurus chrysurus and Scarus
iserti displayed a Stepwise PLCM pattern in which smallest juveniles dwell in the mouth of the bay, larger individuals then
move to habitats deeper into the bay, where they grow up to a (sub-) adult size at which they migrate to nearby coral reef
habitats. This type of stepwise migration in opposite directions, combined with different preference for either mangroves
or seagrass beds among (size-classes of) species, shows that reef fish using in-bay habitats during post-settlement life
stages may do so by choice and not merely because of stochastic dispersal of their larvae, and underline the necessity of
2002 Elsevier Science Ltd. All rights reserved.
these habitats to Caribbean coral reef systems.
Keywords: fish; nursery grounds; mangrove swamps; seagrass; coral reef; Caribbean Sea; migration; Curacao
¸
Introduction efficiency, lower predator abundance, and high inter-
ception rate of the vegetation to planktonic larvae
In many studies, juveniles of reef fish species were
(Odum & Heald, 1972; Carr & Adams, 1973; Ogden
found in high densities in non-reef habitats, while the
& Ziemann, 1977; Blaber & Blaber, 1980; Shulman,
adults were found almost exclusively on the coral
1985; Parrish, 1989; Heck & Crowder, 1991;
reef itself (Pollard, 1984; Parrish, 1989). From this
Robertson & Blaber, 1992).
spatially heterogeneous size-frequency distribution,
Most authors, however, have focused on one or
Post-settlement Life Cycle Migration (PLCM) pat-
two habitats of the mangrove-seagrass-reef con-
terns were suggested that gave birth to the nursery
tinuum, often with different sampling methods, thus
concept. Mangroves and seagrass beds are considered
complicating comparisons among studies and among
nurseries to some reef fish species in the Western
habitats (e.g. Robertson & Duke, 1987; Thayer et al.,
Atlantic, Indian Ocean and Pacific Ocean (Pollard,
1987; Yanez-Arancibia et al., 1988; Blaber et al.,
´˜
1984; Parrish, 1989). Several authors have suggested
1989; Baelde, 1990; Rooker & Dennis, 1991;
the benefits of nurseries to juvenile reef fish, vary-
Sedberry & Carter, 1993; Laegdsgaard & Johnson,
ing from high food availability to lower predation
2001). Quantitative data on ontogenetic shifts in
habitat use from nursery to adult reef association are
b
Corresponding author. E-mail: gerardv@sci.kun.nl
2002 Elsevier Science Ltd. All rights reserved.
0272–7714/01/080309+13 $35.00/0
310 E. Cocheret de la Moriniere et al.
`
largely lacking (Ogden & Ehrlich, 1977; Weinstein & Methods
Heck, 1979; Rooker & Dennis, 1991; Appeldoorn
et al., 1997; Nagelkerken et al., 2000a) and the relative Study area
importance of these nurseries to different size-classes
The present study was carried out in Spanish Water
of reef fish species is still poorly known (Ogden &
Bay in Curacao, Netherlands Antilles (Figure 1).
¸
Gladfelter, 1983; Birkeland, 1985).
This 3 km2 bay is shallow (largely <6 m deep), har-
In recent underwater visual surveys in seven
bours extensive seagrass meadows and is fringed by
different habitats in a marine island bay in Curacao ¸
mangroves (Rhizophora mangle). Water depths under
(Netherlands Antilles), a number of reef fish species of
mangrove canopies ranged between 0·8 m and 1·8 m.
which juveniles were highly abundant in bay environ-
These canopies provide dark habitats (average light
ments were identified and grouped as ‘ nursery
extinction underwater was 85%, as opposed to 40%
species ’ (Nagelkerken et al., 2000a). Nagelkerken
over seagrass beds). The seagrass beds are dominated
et al. (2000a) used a low frequency of surveys in a
by monospecific stands of Thalassia testudinum
large number of transects, and focused on fish com-
(Kuenen & Debrot, 1995). Mean shoot density
munity structure in a range of habitats (mangroves,
( SD) in the seagrass transects was 246 m2 ( 110)
seagrass beds, algal beds, channel, fossil reef terrace
and seagrass canopy height averaged 28·0 cm
notches, boulders, coral reef). Of all habitats in that
( 11·5).
study, seagrass meadows proved to contain highest
There is no freshwater input into the bay other than
total numbers of fish, calculated from observed
rain, and salinity (avg. 35·4) is slightly higher than on
density and total surface area.
the reef (avg. 34·6). Bay water temperature averaged
In the same clear water marine bay in the
30·1 C ( 0·8), while water temperature on the reef
Caribbean, a selection of seven ‘ nursery species ’ of
averaged 28·4 C ( 0·9). Visibility was high at all
which juvenile individuals had been found in large
sites, and varied between an average of 6·5 m ( 1·8)
numbers in mangroves and seagrass beds was
in the bay and 21·4 m ( 3·1) on the reef as measured
studied in detail (Acanthurus chirurgus, Haemulon
by means of a horizontal Secchi disk. The average
flavolineatum, H. sciurus, Lutjanus apodus, L. griseus,
tidal amplitude in the area is 30 cm (De Haan &
Ocyurus chrysurus, and Scarus iserti). In addition to this
Zaneveld, 1959).
set of herbivorous, zoobenthivorous and piscivorous
The bay has a long (1 km) and narrow ( 70 m)
fish, two congeneric species were selected that
entrance that connects it to the adjacent fringing reef.
were encountered in significant quantities in some
This reef is part of a marine park that stretches up to
seagrass beds near the adjacent fringing reef (Scarus
the southwest tip of the island. The reef system starts
taeniopterus and Acanthurus bahianus). Using daytime
with a shallow reef flat (from 2–7 m depth), typically
underwater visual census as a single method to
covered by gorgonians, at the edge of which the
quantify the abundance of the nine selected species
drop-off is located (at 5–10 m depth). Coral cover on
and estimate their size, heterogeneity of the spatial
the drop-off and reef wall is predominated by the
size-frequency distribution of these fish species in reef
stony coral Montastrea annularis. A detailed descrip-
habitats, mangroves and seagrass meadows was
tion of the reefs in the Netherlands Antilles can be
tested. In this way, association of specific size-classes
found in Bak (1975).
of reef fish with specific habitats or spatially separated
sites provides information from which Post-settlement
Life Cycle Migration (PLCM) patterns can be
Sampling design
derived, taking day-to-day variation over a five month
period into account. Additionally, differences in A total of 35 permanent transects were used in 11 sites
spatial distribution and habitat preference can be on the reef and in the bay (Table 1), covering a total
area of about 4500 m2. Each of the transects was
compared among species.
The questions that will be addressed are: censused 29 times on average, during daytime in May
1. Do size-classes of the selected species display any through to September 1998. In Spanish Water Bay,
preference for mangrove or seagrass habitats in six seagrass sites were selected (Figure 1). At each
terms of densities? seagrass site, three permanent 3 by 50 m belt transects
2. Do habitats differ in the size-structure of the were placed, which were surveyed by snorkelling.
subpopulations that they harbour? Average water depth of these transects was between
3. Which spatial migration patterns can be inferred 0·8 and 2·4 m. Adjacent to four of the seagrass sites
from average densities and sizes by comparison of (numbered 2, 3, 4 and 6) was a mangrove site (Figure
the subpopulations at the various sites? 1). The mangrove stands consist of strips of vegetation
Migration patterns and habitat preference of coral reef fishes 311
Curaçao
N
25°
N
20°
Trade wind
Caribbean Sea
15°
10°
Spanish Water Bay
90° 85° 80° 75° 70° 65° 60°
5m
6
5
m
5m
5m
5 5m
3
10 m
10 m
4
10 m
5m
2
Mangroves
0 250 500 1
Seagrass beds
m
Coral reef
7
Spanish Water Bay
F 1. Location of the study sites in Spanish Water Bay. At sites 1–6 the seagrass beds were censused, while at site 2, 3,
4, and 6 mangroves were also surveyed. Site 7 was the reef site.
hanging over from fossil reef ledges, hence providing (Table 1). At the reef site (numbered 7, Figure 1),
structural complexity from prop roots or branches in three permanent 3 by 50 m belt transects were placed
the water column beneath the mangrove canopy. The at three depths (5, 10 and 15 m) parallel to the
eight mangrove transects were narrow underwater coastline, using nylon twine. The three 5 m deep
habitats, and were censused by snorkelers. In the transects were placed where the sandy reef flat ends at
mangroves, transect width was between 1·1 and 2·1 m the start of the drop-off, the three 10 m deep transects
312 E. Cocheret de la Moriniere et al.
`
T 1. Surface area of the 35 transects in all sites and transects was censused 29 times on average). Data
habitats. Site numbers correspond to Figure 1. On the coral were logtransformed and analysed in a nested
reef, 3 transects were used at each depth, as indicated
ANOVA (GLM, SPSS 8.0) for unequal sample sizes,
between brackets
where sites were nested in habitats and individual
surveys of the transects were treated as replicates
Width Length Area
(m2) within sites. Multiple comparisons of means within
Site Habitat Transect (m) (m)
habitats (among sites) and among habitats were
analysed using a Tukey HSD Spjotvoll/Stoline test
1–6 Seagrass 1 3·0 50·0 150·0
(Sokal & Rohlf, 1995). Observer effect was tested by a
2 3·0 50·0 150·0
3 3·0 50·0 150·0 one-way ANOVA on each of the 35 transects, with
observer identity as an independent variable and mean
2 Mangrove 1 1·3 37·5 48·8
2 1·3 40·0 52·0 size or total density in the surveys as a dependent
3 1 1·1 73·0 80·3 variable. None of the 35 ANOVAs on observer effect
4 1 1·4 43·0 60·2
produced significant differences (P<0·05) in variance
2 1·3 10·0 13·0
among size or density estimation among observers.
6 1 2·1 38·0 79·8
Since there is a variety of prevailing habitats in
2 1·4 37·0 51·8
3 1·5 31·0 46·5 Spanish Water Bay (mangroves, seagrass beds, algal
beds, channel, fossil reef terrace notches, boulders,
7 Reef 1 (3 at 5 m depth) 3·0 50·0 150·0
2 (3 at 10 m depth) 3·0 50·0 150·0 coral reef; see Nagelkerken et al., 2000a), and the
3 (3 at 15 m depth) 3·0 50·0 150·0 selected species use these shallow habitats as daily
resting sites to which they return every day after
nocturnal migrations to deeper feeding or sleeping
grounds (Nagelkerken et al., 2000b), their daytime
were located on the drop-off whereas the three 15 m density distribution can be viewed as a matter of
deep transects were situated on the reef slope. At each choice. Therefore, the density of a size-class of a fish
depth, the three transects were placed 50 m apart species in mangroves relative to its density in seagrass
from each other. The depth range was based on a beds is viewed as a level of habitat preference. The
pilot-study that showed that the selected species level of preference for either mangroves or seagrass
reached highest densities at depths less than 15 m. beds was tested based on densities of the size-classes
Reef sites were censused by Scuba diving. of each fish species occurring in mangroves and
During visual surveys, individuals of the selected seagrass beds at site numbers 2, 3, 4 and 6 (Figure 1).
species were counted and their sizes estimated in Only these sites were used for analysis of habitat
size-classes of 2·5 cm. Underwater size estimators preference because both seagrass and mangrove
were trained with objects of known size. The three habitats were surveyed at those sites. For each size-
observers censused all transects using an alternating class of each species, the average density in mangroves
system so that any bias in size-estimation is equally at a site was divided by the sum density of that
represented in every transect. The observer effect was size-class in mangroves and seagrass beds at that site.
tested using ANOVA (see ‘ Statistical analysis ’ for These mangrove-to-seagrass preference levels of the
further explanation). All juvenile fishes observed in size-classes of the species at the four sites were then
this study were larger than 1 cm at settlement. Juv- clustered using City-block (Manhattan) distances
enile scarids smaller than 5 cm (TL) could not be (Statistica for Windows 4·5). In the Manhattan dis-
identified in the field. Scarids of sizes smaller than tances measure, the effect of single large differences
5 cm were left out of the data sets of Scarus iserti and (outliers) is dampened (in the Euclidean distance
S. taeniopterus. Juveniles and sub-adults of these two measure, differences are squared).
scarids that were larger than 5 cm (TL) could be
distinguished by the characteristics shown in Humann
Results
(1996). All other species could be identified at all
sizes.
Habitat preference
Mean densities (100 m 2) of most species are signifi-
Statistical analysis
cantly lower on the coral reef than in seagrass or
For each species, mean size (cm) and total density mangrove habitats, with the exception of Ocyurus
(N 100 m 2) of the observed individuals was calcu- chrysurus (coral reef densities similar to densities in
lated at each survey of a transect (each of the 35 seagrass beds and lowest densities in mangroves) and
Migration patterns and habitat preference of coral reef fishes 313
T 2. Average size (cm) and density (N 100 2) per species in each habitat, and their standard errors between brackets.
Among sizes, significant (P<0·05) differences are indicated with a, b, and c for each species. Among densities, significantly
different means are marked d, e, or f for each species. Different letters (a–f) mean that averages are significantly different
Size Densities
Mangrove Seagrass Reef Mangrove Seagrass Reef
4·4a (0·4) 12·9b (0·6) 4·3d (0·5) 4·9d (1·6)
Acanthurus bahianus — 0·0
13·0a (0·3) 11·0b (0·3) 17·0c (0·5) 3·3d (0·3) 0·9e (0·2) 2·8f (0·9)
Acanthurus chirurgus
8·8a (0·2) 7·7b (0·1) 15·1c (0·2) 99·8d (12·2) 32·1e (1·7) 3·2f (0·4)
Haemulon flavolineatum
12·3a (0·2) 11·5b (0·1) 21·5c (0·3) 18·2d (1·2) 5·7e (0·3) 0·3f (0·0)
Haemulon sciurus
12·3a (0·2) 11·3a (0·5) 18·5b (0·5) 24·7d (1·4) 0·3e (0·1) 1·0f (0·1)
Lutjanus apodus
14·2a (0·3) 12·6b (0·3) 16·6a (1·1) 8·1d (0·8) 0·5e (0·1) 0·0f (0·0)
Lutjanus griseus
9·3a (0·4) 9·8a (0·2) 17·1b (0·3) 2·1d (0·4) 3·6e (0·4) 4·1e (0·5)
Ocyurus chrysurus
8·1a (0·3) 7·8a (0·1) 11·9b (0·3) 11·5d (2·0) 13·3d (1·3) 2·9e (0·2)
Scarus iserti
6·3a (0·0) 6·4a (0·1) 16·4b (0·3) 6·3d (1·3) 1·7e (0·3) 3·6f (0·2)
Scarus taeniopterus
Acanthurus bahianus (not observed in mangroves, and ever-larger sizes found in clusters with an increasing
seagrass densities not significantly different from reef mangrove-to-seagrass density ratio. The latter two
densities) (Table 2). Overall densities of Scarus iserti, species suggest that their smallest juveniles are most
Ocyurus chrysurus, and Acanthurus bahianus are higher commonly found in seagrass beds, while mangrove
in seagrass beds than in mangroves, while the reverse preference increases with size of the juveniles. Only
is true for the remaining species (Table 2). one size-class (5·0–7·5 cm) of Scarus taeniopterus was
In order to determine the level of habitat observed in the mouth of the bay, of which indi-
preference, mangrove-to-seagrass density ratios (see viduals were observed in mangroves and seagrass beds
Statistical analysis) were determined for each size-class (cluster B).
at four sites where both mangroves and seagrass beds
were surveyed. Cluster analysis of these mangrove-to-
Size-distribution over habitats
seagrass density ratios at the four sites (numbered 2,
3, 4, and 6 in Figure 1) yielded three distinct groups of In Figure 3, relative densities of the selected species in
size-classes of fishes at linkage distance 2·0 (Figure 2). the three habitats are depicted for each size-class.
Cluster A had an average mangrove-to-seagrass All selected species were ‘ nursery species ’, in the
density ratio of 18% (range 0–46%); in cluster B, that sense that high densities of juveniles were found in
ratio is 60% (34–75%), and in cluster C it is 93% mangroves or seagrass beds, while most adults were
(64–100%). All size-classes of Lutjanus apodus and observed on the reef. All three habitats differed
Haemulon sciurus belonged to cluster C, reflecting a significantly in the average sizes of individuals of
strong preference for mangroves at all size-classes. Haemulon flavolineatum, H. sciurus, and Acanthurus
Lutjanus griseus also seems to prefer mangroves over chirurgus that they harboured (Table 2). Ocyurus
seagrass beds, since all size-classes were members of chrysurus, Lutjanus apodus, Scarus iserti, and S.
cluster C, and one size-class (7·5–10 cm) was in taeniopterus showed no difference in average sizes
cluster B. Acanthurus bahianus was only observed in between mangroves and seagrass beds, but individuals
seagrass habitats, and therefore all size-classes were on the reef were significantly larger. Acanthurus
members of cluster A. Acanthurus chirurgus distributed bahianus was never found in mangroves, and the
itself over mangroves and seagrass beds (most size- average size of the individuals of this species observed
classes are members of cluster B), with a few individ- in seagrass beds was smaller than on the reef. Average
uals of the largest size-class found only in mangroves size of Lutjanus griseus was significantly smaller in
(and therefore part of cluster C). Ocyurus chrysurus seagrass beds than in mangroves and on the coral reef,
was observed in both seagrass beds and mangroves while the latter two habitats showed average sizes that
(size-classes are part of clusters A and B). Scarus iserti were similar to each other. Mean size of all nine
occurred mostly in seagrass beds (cluster A), but species on the reef (Table 2) was always smaller than
mangrove preference seems to increase with size or corresponded to the approximate mean total
(largest sizes in cluster B). Haemulon flavolineatum is lengths at which these species become sexually mature
represented by size-classes in all three clusters, with (see Figure 3).
314 E. Cocheret de la Moriniere et al.
`
Linkage distance
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
O. chrys 17.5–20
O. chrys 2.5–5
O. chrys 15–17.5
A. bah 12.5–15
A. bah 10–12.5
A. bah 7.5–10
A. bah 5–7.5
A. bah 2.5–5 A
A. bah 0–2.5
S. iser 7.5–10
S. iser 5–7.5
O. chrys 5–7.5
O. chrys 7.5–10
S. iser 10–12.5
H. flav 5–7.5
H. flav 2.5–5
H. flav 0–2.5
O. chrys 20–22.5
O. chrys 0–2.5
S. taen 5–7.5
O. chrys 12.5–15
B
O. chrys 10–12.5
H. flav 10–12.5
H. flav 7.5–10
A. chir 10–12.5
L. gris 7.5–10
S. iser 12.5–15
A. chir 15–17.5
A. chir 7.5–10
A. chir 12.5–15
S. iser 15–17.5
A. chir 20–22.5
A. chir 5–7.5
A. chir 2.5–5
A. chir 17.5–20
A. chir 0–2.5
L. apo 5–7.5
L. apo 2.5–5
L. apo 7.5–10
L. gris 15–17.5
L. apo 12.5–15
L. apo 10–12.5
L. apo 15–17.5
L. gris 17.5–20
L. apo 17.5–20
H. flav 15–17.5
L. apo 25–27.5
L. apo 22.5–25
L. apo 20–22.5
L. apo 0–2.5
L. gris 20–22.5
A. chir 22.5–25
H. sci 17.5–20
L. gris 22.5–25
H. sci 20–22.5
L. gris 12.5–15
L. gris 10–12.5 C
L. gris 5–7.5
H. flav 12.5–15
H. sci 15–17.5
H. sci 12.5–15
H. sci 10–12.5
H. sci 7.5–10
H. sci 5–7.5
H. sci 2.5–5
H. sci 0–2.5
F 2. Complete linkage of relative (mangrove-to-seagrass) densities of the size-classes of the study species in four sites
(numbered 2, 3, 4 and 6; see Figure 1), using City-block (Manhattan) distances. Species are indicated by the following codes:
A. bah=Acanthurus bahianus, A. chir=Acanthurus chirurgus, H. flav=Haemulon flavolineatum, H. sci=Haemulon sciurus,
L. apo=Lutjanus apodus, L. gris=Lutjanus griseus, O. chrys=Ocyurus chrysurus, S. iser=Scarus iserti, S. taen=Scarus taeni-
opterus. Size-classes (cm) are indicated by the numbers behind the species codes.
Migration patterns and habitat preference of coral reef fishes 315
Spatial migration patterns of a fish species over mangroves and seagrass beds at
daytime is considered here as a matter of choice.
The size-frequency distribution of Haemulon
Cluster-analysis of the mangrove-to-seagrass density
flavolineatum, H. sciurus, Acanthurus chirurgus,
ratios of each size-class, at four sites that harboured
Lutjanus apodus, Scarus iserti, and Ocyurus chrysurus in
both habitats showed different levels of habitat pref-
mangroves and seagrass beds and on the reef suggests
erence. Lutjanus apodus and Haemulon sciurus showed
a size-range for each species over which the juveniles
strong preference for mangroves over seagrass beds at
start migrating to the coral reef (Figure 3). In the case
all size-classes. L. griseus was also strongly associated
of Scarus taeniopterus this migration from nursery
with mangroves at all size-classes, and moderately by
habitat to coral reef appears to take place rather
one size-class. Acanthurus bahianus was not observed
abruptly, while individuals of Acanthurus bahianus
in mangroves, reflecting strong preference for seagrass
may migrate to reef habitats at all sizes.
beds. Ocyurus chrysurus and Acanthurus chirurgus
When size-frequencies were compared among bay
utilized both habitats. Scarus iserti and Haemulon
sites within each habitat (ANOVA), heterogeneous
flavolineatum also used both habitats, but there was a
distribution patterns over sites emerged for some
trend of increased preference for mangroves with
species (Table 3). For H. sciurus, the extremely low
increasing fish size, while the smallest juveniles of
(0·09 100 m 2) densities of small individuals at site 2
these species were highly associated with seagrass
are responsible for significant differences among
beds. Though seemingly marginal habitats, strips of
seagrass sites.
mangroves of no more than 1 by 40 m at times
Haemulon flavolineatum, Scarus iserti, and Ocyurus
may contain hundreds of individuals in resting
chrysurus displayed a size-frequency distribution in
schools. The preference of Scarus taeniopterus for
which high densities of small individuals were found
mangroves may be exaggerated (mangrove-to-seagrass
in mangroves and seagrass beds in the mouth of the
density ratio was about 60% : 40%), since juveniles
bay (at site numbers 1 and 2, Figure 1), medium-sized
smaller than 5 cm were excluded from the data set.
fishes deeper in the bay, and large fishes on the reef
Unidentifiable scarid juveniles of this size were mostly
[Figure 4(a, b and c)]. This indicates a Post-
found in the mouth of the bay (average density
settlement Life Cycle Migration (PLCM) pattern with
38·2 100 m 2 in the mouth of the bay as opposed to
two changes of direction.
0·8 100 m 2 in transects deeper in the bay) in sea-
Both Acanthurus bahianus and Scarus taeniopterus
grass beds. Nagelkerken et al. (2000a, c) have found
were only encountered at sites 1 and 2, which are
similar overall density distributions of these species in
located in the mouth of the bay (see Figure 1), and on
mangrove habitats and seagrass habitats in Curacao ¸
the reef. They were not observed at sites located
and Bonaire. The level of preference of these fish
deeper into the bay. In both cases, high densities of
species for mangroves or seagrass beds in the situation
small juveniles were detected in the mouth of the bay
where both habitats occur, however, is no indication
[Figure 4(d and e)], while statistically lower densities
of the level of dependence on these habitats. From
and larger individuals of these species occurred on the
comparisons among bays with and without mangroves
reef (Table 2). Their distribution indicates a PLCM
or seagrass beds (Nagelkerken et al., 2001), it is
pattern that is restricted to seagrass and mangrove
known that species that showed strong preference for
sites in the close vicinity of the reef, and does not
mangroves in the present study (Lutjanus apodus, L.
include temporary residence deeper in the bay.
griseus, Haemulon sciurus) depend largely on the pres-
ence or absence of seagrass beds. Given the choice,
such species apparently prefer mangroves as daytime
Discussion
resting sites for shelter, while their dependence on
seagrass beds is best explained by the larger abun-
Habitat preference
dance of food in seagrass habitats in which they forage
From previous studies in the same bay (Nagelkerken at night.
et al., 2000a, b), it is known that fish have a number of
occurring habitats to choose from in Spanish Water
Size-distribution and spatial migration patterns
Bay (mangroves, seagrass beds, algal beds, channel,
fossil reef terrace notches, boulders). These shallow All selected species proved to be ‘ nursery species ’ in
habitats are used as daily resting sites, to which the the sense that juveniles were much more abundant in
fishes return every day after nocturnal migrations to mangroves or seagrass beds than on the reef, as
deeper feeding or sleeping grounds (Nagelkerken et expected from our previous study (Nagelkerken et al.,
al., 2000b). Therefore, the relative density distribution 2000a). Of nine species, six (Haemulon flavolineatum,
Average density (%) Average density (%)
Average density (%) Average density (%) Average density (%)
0. 0.
5. 0. 0.
0
0
0–
0
20
40
60
80
100
0– 0–
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
–7
2. 2. –2. 2. 2.
55
.5 5 5
2.
2. 5 2.
5–
5– 5. –5.
7. 5–
00 5.
5. 5–
0 5.
5.
7. –7 0
5.
10
5. 0 0– 5.
.0
0– 10 –1 5 7. 0–
7.
.0
10
7.
5 5– 5 7.
.0 5
7. 12 0–1 .0 7.
Seagrass
.2 10 10
–1
5–
Seagrass
Mangrove
5–
.0 .0
2.
10 15 5–1 .5 –
5 10
Seagrass
.0
10 12 .0 5. .0
Seagrass
12 12 10
.0 .5 17 –1 0
Mangrove
Mangrove
.5 .5 .0
.7 –
–1 –1
`
Seagrass
–1
20 5–2 .5
5. 15 15
12 2.5
0 2.
12 5
.0 .0
.5 .0 0.
15
– .5
22 –2 0
.0
–1
.2 –1
17 17
–1
Mangrove
15 5.0 .5 .5 5.
25 5–2 .5 15 0
7. –
.0 .5
5 .0
316 E. Cocheret de la Moriniere et al.
17
–1 20 20
27 0–2 .0
(e) L. apodus
–1
(i) A. bahianus
.5 .0 .0
(c) A. chirurgus
Size-classes (cm)
17 7.5
Reef
(g) Scarus iserti
–
.5 7. 7.
17
–2 5
.5 30 –3 5 .5
22 22
(a) H. flavolineatum
.0
0.
–2 0 .5 .5 –2
20
–
Reef
32 0–3 .0 Reef
20 0.0
.2 0.
.0 20 0
.0 25 25 .0
–2 35 5–3 .5 .0 .0
–2
.5 – –2
2.
Reef
5
22 2.5 2.
Reef
37 0–3 .0 27 27
22 22 5
.5 .5 .5
.5 .5
.5 7.
–2 –4 5 –3
–2 –2
5.
0 0. 0.
5. 5.
0 0
0 0
Average density (%) Average density (%)
Average density (%) Average density (%)
0. 0.
5. 5.
0– 0–
0– 0–
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
2.
7. 7.
2. 2.5 5
5 5 2.
7.
5– 7. 5–
5–
5–
5. 5.0 5.
0 0
10
10 5.
.0
10
.0 0–
7. –7. 10 .0
5– 5 7.
7.
.0 –1 5– 5
10 10 –1 2.
12
.0 .0 5
Mangrove
Seagrass
10 10
2. .5
5
12 .0 .0
12 –1
Seagrass
–
–1
.5
.5 2.
Seagrass
5.
–1 12 12
15 0
15 –1 5 .5 .5
Mangrove
.0
5.
.0 5. 0 –
15 –1
.0 15 15
17 –1 0
Seagrass
7.
17 .0 .0
–1
.5 7.5 .5 5 –1
7.
20 –2 5 –2
17 17 7
.0 0. .5 .5 .5
20 0.0
–
22 –2 0 –2
Mangrove
.0
.5 2. 20 20
(f) L. griseus
0. –2
Mangrove
(b) H. sciurus
0 .0 .0
20
25 –2 5 –
Size-classes (cm)
.0 22 2.5
Reef
.0 5. .5 22 22
Reef
–2
(h) Ocyurus chrysurus
.5 .5
27 –2 0 –2
(d) Scarus taeniopterus
2.
–
5
.5 7. 22 25 5.0
.5 25 25
.0
30 –3 5 .0 .0
–2
Reef
.0 0. –
–2
5.
0 27 27
32 –3 0 25
Reef
27 7.5
.5 .5
.0 .5
.5 2.5
–3
–3 –2 –3
0.
5. 7. 0.
0
0 5 0
Migration patterns and habitat preference of coral reef fishes 317
T 3. Analysis of variance of sizes among sites within bay assumed that it is a ‘ nursery species ’ as well, and
habitats. Bold print represents P-values smaller than 0·05. that it migrates according to the above mentioned
ANOVA could not be applied to A. bahianus and S.
Long Distance PLCM pattern. This distribution
taeniopterus. Significant P-values indicate that variance of
pattern fits the observed size-distribution by other
sizes of fishes is not homogeneously distributed over sites
authors well (e.g. Parrish, 1989 and references
within that habitat
therein; Rutherford et al., 1989; Appeldoorn et al.,
Mangrove Seagrass 1997), with larger individuals found progressively
off-shore.
Three species, Haemulon flavolineatum, Ocyurus
Acanthurus bahianus — 0·44
Acanthurus chirurgus 0·20 0·16 chrysurus and Scarus iserti, showed similar size-
Haemulon flavolineatum 0·00 0·00 frequency distribution patterns with three spatially
Haemulon sciurus 0·20 0·00
separated groups that were statistically different: small
Lutjanus apodus 0·15 0·16
juveniles in the mouth of the bay, larger individuals
Lutjanus griseus 0·35 0·27
in seagrass beds and mangroves located deeper into
Ocyurus chrysurus 0·10 0·00
Scarus iserti 0·00 0·00 the bay and (sub-) adults on the reef. This spatial
Scarus taeniopterus — 0·00 distribution suggests a Life Cycle Migration that
involves more than one direction of migration, and is
hence termed Stepwise PLCM. Post-larvae of these
species settle in the mouth of the bay, after which they
H. sciurus, Lutjanus apodus, Ocyurus chrysurus, migrate deeper into the bay to grow up to a size large
Acanthurus chirurgus, and Scarus iserti) showed enough to migrate to and dwell on the reef. Most
spatial distributions in which smallest individuals migrants into coastal regions come from the open sea
were only found in bay habitats. The largest indi- (Blaber, 1997). These are in-out migrations: juveniles
viduals of Haemulon flavolineatum, Lutjanus apodus, or adults or both migrate into an estuarine or coastal
Ocyurus chrysurus, and Scarus iserti were found only on area for a certain period, after which they return to the
the adjacent reef, while adults of Haemulon sciurus, open sea or coral reef. De Sylva (1963) describes a
and Acanthurus chirurgus were found both in reef distribution pattern of Sphyraena barracuda that is
habitats as well as in bay habitats (mangrove or coherent to the Stepwise PLCM patterns that were
seagrass). Average size of these six species was largest found in our study. Post-larvae and juveniles of this
on the reef. The size at which these species become piscivore move from coastal shallows to reed beds or
sexually mature (Robertson & Warner, 1978; Munro, mangroves, followed by a migration to open sea.
1983; Munro pers. comm.) always corresponded to or Juveniles of Scarus taeniopterus and Acanthurus
was larger than the average size at which they were bahianus were only found in the mouth of the bay and
found on the coral reef. The results suggest a Post- on the reef. Small juveniles of Acanthurus bahianus
settlement Life Cycle Migration (PLCM) pattern over were only observed in seagrass beds in the mouth of
a considerable distance, in which juveniles settle the bay and in the reef flats, while adults occurred
and grow up in alternative habitats such as seagrass almost exclusively on the reef. These two species
beds and mangroves, after which the sub-adults display a type of Short Distance PLCM in which
migrate to reef habitats where they become sexually larvae partly settle in the mouth of the bay and partly
mature. That pattern is named Long Distance in the reef flats, to reach a size at which they migrate to
PLCM. Average size of Lutjanus griseus was smallest in deeper reef habitats.
seagrass beds, while average sizes in mangroves and Interestingly, some congeneric species appeared to
on reefs were similar. The size at which Lutjanus display different directions of migration at similar
griseus becomes sexually mature is about 25 cm sizes. Scarids of similar sizes are found mixed in the
(Starck, 1971; Claro, 1983), and individuals of this mouth of the bay, but one species then migrates to
size have been observed in mangroves as well as on reef habitats (Scarus taeniopterus), and the other
the reef. Since small juveniles of Lutjanus griseus migrates deeper into the bay and only dwells on the
were only observed in bay habitats, and spawning reef at larger sizes (Scarus iserti). Leaving unidenti-
occurs on the reef or shelf edge (Claro, 1983), it is fiable scarids smaller than 5 cm out of the data sets
F 3. Average relative density of individuals of the selected species in three habitats, calculated per size-class. The
arrows indicate approximate length at sexual maturity (maturity data from Munro, 1983; Starck, 1971; Claro, 1983).
318 E. Cocheret de la Moriniere et al.
`
18 17
(a) (b)
16
15
14
Mean size (cm)
13
12
d
cd c
10 11
ce e c
b
b b
a b
8
b 9
6 a
b
a b
a a
7
4
a
2 5
M2 M3 M4 M6 S1 S2 S3 S4 S5 S6 Reef M2 M4 M6 S1 S2 S3 S4 S5 S6 Reef
Sites Sites
24 24
(c) (d)
20 20
Mean size (cm)
16 16
bc
b cd
d
12 12
a
b
a
b
8 8
a a b
ab
4 4
0 0
M2 M3 M4 M6 S1 S2 S3 S4 S5 S6 Reef M2 S1 S2 Reef
Sites Sites
20
(e)
18
16
Mean size (cm)
14
12 ± Std. Dev.
10 Mean
8
a
a
6
4
2
S1 S2 Reef
Sites
F 4. Mean and standard deviation of size-distribution of Haemulon flavolineatum (a), Scarus iserti (b), Ocyurus chrysurus
(c) Scarus taeniopterus (d) and Acanthurus bahianus (e) at each site. Within each habitat (indicated with ‘ S ’, ‘ M ’, or ‘ Reef ’,
for seagrass, mangrove and reef habitats, respectively), statistically significant differences among sites (Tukey HSD) are
presented by letters only when the species was found in that habitat at more than one site. Different letters mean that the
averages of the sites are statistically different.
did not affect conclusions regarding the spatial distribution, abundance and early post-settlement
migration patterns of these species, since this size- persistence of settlers (Risk, 1998), while post-
class was almost exclusively found in the mouth of settlement habitat selection is important in creating
the bay. Apparently, both scarids settle in seagrass spatial patterns of recruitment (Sponaugle & Cowen,
beds and mangroves located in the mouth of the bay, 1996). This means that competitive congeneric
after which each migrates in an opposite direction. species can alleviate competition on the reef by
The same difference is observed when comparing the temporary spatial separation.
migration patterns of Acanthurus chirurgus and A. Possible explanations for different spatial size-
bahianus. Of Acanthurus bahianus it is known that frequency distributions of post-settlement fishes
behavioural interactions are size-related and can affect involve variability in mortality rates, growth,
Migration patterns and habitat preference of coral reef fishes 319
F 5. Three types of migration among nurseries and the coral reef. Route 1 depicts the Short Distance PLCM, route 2
Stepwise PLCM, and route 3 Long Distance PLCM (explanation see text).
settlement patterns and migration patterns. Since the Conclusions
abundance of predators is much lower in the bay
than on the reef as is generally the case (Shulman, Taking day-to-day variations in fish density and size-
1985; Parrish, 1989), differences in mortality rates frequency into account over a five-month period,
may explain the high abundance of juveniles in bay spatial patterns emerge for the selected fish species. Of
habitats and the reduced numbers on the coral reef. some species, all size-classes that occurred in bay
In fact, reduced mortality among juveniles in nursery habitats appeared to prefer mangroves as daytime
habitats is often ascribed to reduced predator abun- resting sites, while others were only found in seagrass
dance or efficiency (e.g. Heck & Crowder, 1991; beds. Other species utilized and preferred mangroves
Robertson & Blaber, 1992). This, however, cannot and seagrass beds at different sizes, and preference for
explain the lower number of (sub-)adults in man- mangroves of some species increased with increasing
grove and seagrass habitats or the low numbers of size-class.
the smallest juveniles in the habitats that are located The size-frequency distribution patterns of
deep in the bay. Also, abundance and availability of Haemulon flavolineatum, H. sciurus, Acanthurus
food items (such as benthic and planktonic inverte- chirurgus, Lutjanus apodus, Scarus iserti, and Ocyurus
brates, epifauna and epiphytes) is much higher in the chrysurus in mangroves and seagrass beds and on
bay habitats of Spanish Water Bay than on the the reef suggest a size-range for each species over
nearby coral reef (Cocheret de la Moriniere et al.,
` which the juveniles start migrating to the coral reef
unpublished), which could not result in lower (Figure 3). In the case of Scarus taeniopterus this
growth rates of fishes in the bay. Variability in migration from nursery habitat to coral reef appears to
growth rate is therefore another unlikely explanatory take place rather abruptly, while individuals of Acan-
factor for the fact that the largest individuals of thurus bahianus may migrate to reef habitats at all
nursery species are usually found on the coral reef. sizes.
The spawning seasons of the selected species are Haemulon sciurus, Lutjanus apodus, L. griseus, and
largely during the study period (Munro et al., 1973), Acanthurus chirurgus display Long Distance PLCM;
and regular settlement (no major peaks) of post- Haemulon flavolineatum, Ocyurus chrysurus and
larvae was observed for most of the species during Scarus iserti use Stepwise PLCM; Scarus taeniopterus
the study. Considering all these processes, migration and Acanthurus bahianus are retained within a small
from nursery ground to coral reef habitat seems a distance from the reef (Short Distance PLCM).
logical explanation for the spatial distribution of These different migration patterns are depicted in
size-classes of these fish species, and migratory pat- Figure 5. The fact that some species carry out specific
terns can be inferred. The stability of such patterns directional migrations and congeners may migrate
and validity of actual migrations must be tested in to different areas raises questions concerning the
further studies. mechanisms that trigger these migrational options,
320 E. Cocheret de la Moriniere et al.
`
Heck, K. L. & Crowder, L. B. 1991 Habitat structure and predator-
and their ecological or evolutionary meaning. This
prey interactions in vegetated aquatic ecosystems. In Habitat
also encourages scientists to view the ‘ nursery ques- Structure: The Physical Arrangements of Objects in Space (Bell, S. S.,
tion ’ not only from the point of view of benevolence McCoy, E. D. & Mushinsky, E. R., eds). Chapman and Hall,
London, pp. 281–299.
of nursery areas, but also to elucidate why a species
Humann, P. 1996 Reef Fish Identification, 2nd edition. Florida
shows a particular migration pattern or why it does Caribbean Bahamas. New World Publications Inc., Jacksonville,
not. USA, 396 pp.
Kuenen, M. M. C. E. & Debrot, A. O. 1995 A quantitative study of
the seagrass and algal meadows of the Spaanse Water, Curacao, ¸
The Netherlands Antilles. Aquatic Botany 51, 291–310.
Acknowledgements Laegdsgaard, P. & Johnson, C. 2001 Why do juvenile fish utilise
mangrove habitats? Journal of Experimental Marine Biology and
We would like to thank all personnel of the Ecology 257, 229–253.
Carmabi Foundation, where the research was carried Munro, J. L., Gaut, V. C., Thompson, R. & Reeson, P. H. 1973
The spawning seasons of Caribbean reef fishes. Journal of Fish
out. The research was funded by the Netherlands Biology 5, 69–84.
Foundation for the Advancement of Tropical Munro, J. L. (ed.) 1983 Caribbean Coral Reef Fishery Resources, 2nd
Research (WOTRO). The Stichting Nijmeegs edition. ICLARM, Philippines, 276 pp.
Nagelkerken, I., Dorenbosch, M., Verberk, W. C. E. P., Cocheret
Universiteitsfonds (SNUF) funded B. J. A. Pollux.
de la Moriniere, E. & van der Velde, G. 2000a Importance of
`
Furthermore, we thank E. Kardinaal for supplying us shallow-water biotopes of a Caribbean bay for juvenile coral reef
with a map of the Spanish Water Bay and the Winkel fishes: patterns in biotope association, community structure
and spatial distribution. Marine Ecology Progress Series 202, 175–
family for the use of their pier. Finally, we would like
192.
to thank Blu Forman for her help with the graph on Nagelkerken, I., Dorenbosch, M., Verberk, W. C. E. P., Cocheret
migration patterns, and P. H. Nienhuis for useful de la Moriniere, E. & van der Velde, G. 2000b Day-night shifts
`
comments on the manuscript. of fishes between shallow-water biotopes of a Caribbean bay,
with emphasis on the nocturnal feeding of Haemulidae and
Lutjanidae. Marine Ecology Progress Series 194, 55–64.
Nagelkerken, I., van der Velde, G., Gorissen, M. W., Meijer, G. J.,
References van ’t Hoff, T. & den Hartog, C. 2000c Importance of man-
groves, seagrass beds and the shallow coral reef as a nursery for
Appeldoorn, R. S., Recksiek, C. W., Hill, R. L., Pagan, F. E. & important coral reef fishes, using a visual census technique.
Dennis, G. D. 1997 Marine protected areas and reef fish move- Estuarine, Coastal and Shelf Science 51, 31–44.
ments: the role of habitat in controlling ontogenetic migration.
Nagelkerken, I., Kleijnen, S., Klop, T., van den Brand, R. A. C. J.,
Proceedings of the 8th International Coral Reef Symposium 2,
Cocheret de la Moriniere, E. & van der Velde, G. 2001 Depen-
`
1917–1922.
dence of Caribbean reef fishes on mangroves and seagrass beds as
Baelde, P. 1990 Differences in the structures of fish assemblages in
nursery habitats: a comparison of fish faunas between bays with
Thalassia testudinum beds in Guadeloupe, French West Indies,
and without mangroves and seagrass beds. Marine Ecology
and their ecological significance. Marine Biology 105, 163–173.
Progress Series 214, 225–235.
Bak, R. P. M. 1975 Ecological aspects of the distribution of reef
Odum, W. E. & Heald, E. J. 1972 Trophic analyses of an estuarine
corals in the Netherlands Antilles. Bijdragen tot de Dierkunde 45,
mangrove community. Bulletin of Marine Science 22, 671–738.
181–190.
Ogden, J. C. & Ehrlich, P. R. 1977 The behaviour of heterotypic
Birkeland, C. 1985 Ecological interactions between tropical coastal
resting schools of juvenile grunts (Pomadasyidae). Marine Biology
ecosystems. UNEP Regional Seas Reports and Studies 73, 1–26.
42, 273–280.
Blaber, S. J. M., Brewer, D. T. & Salini, J. P. 1989 Species
Ogden, J. C. & Gladfelter, E. H. 1983 Coral reefs, seagrass beds,
composition and biomasses of fishes in different habitats of a
and mangroves: their interactions in the coastal zones of the
tropical northern Australian estuary: their occurrence in the
Caribbean. UNESCO Reports in Marine Science 23, 133 pp.
adjoining sea and estuarine dependence. Estuarine, Coastal and
Ogden, J. C. & Zieman, J. C. 1977 Ecological aspects of coral
Shelf Science 29, 509–531.
reef-seagrass bed contacts in the Caribbean. Proceedings of the 3rd
Blaber, S. J. M. (ed.) 1997 Fish and Fisheries of Tropical Estuaries,
International Coral Reef Symposium 1, 377–382.
Volume 22. Chapman & Hall, London, 353 pp.
Parrish, J. D. 1989 Fish communities of interacting shallow-water
Blaber, S. J. M. & Blaber, T. G. 1980 Factors affecting the
habitats in tropical oceanic regions. Marine Ecology Progress Series
distribution of juvenile estuarine and inshore fish. Journal of Fish
58, 143–160.
Biology 17, 143–162.
Pollard, D. A. 1984 A review of ecological studies on seagrass-fish
Carr, W. E. S. & Adams, C. A. 1973 Food habits of juvenile marine
communities, with particular reference to recent studies in
fishes occupying seagrass beds in the estuarine zone near Crystal
Australia. Aquatic Botany 18, 3–42.
River, Florida. Transactions of the American Fisheries Society 102,
Risk, A. 1998 The effects of interactions with reef residents on
511–540.
the settlement and subsequent persistence of ocean surgeon-
Claro, R. 1983 Ecologia y cicle de vida del caballerote,
fish, Acanthurus bahianus. Environmental Biology of Fishes 51,
Lutjanus griseus (Linnaeus), en la platforma cubana. I. Identidad,
377–389.
distribucion y habitat, nutricion y reproduccion. Reportajes y
Investigaciones del Instituto por Oceanologia de la Academia de Robertson, A. I. & Duke, N. C. 1987 Mangroves as nursery sites:
Sciencia de Cuba 7, 1–30. comparisons of the abundance and species composition of fish
and crustaceans in mangroves and other nearshore habitats in
De Haan, D. & Zaneveld, J. S. 1959 Some notes on tides in
tropical Australia. Marine Biology 96, 193–205.
Annabaai harbour, Curacao, Netherlands Antilles. Bulletin of
¸
Marine Science of the Gulf and Caribbean 9, 224–236. Robertson, A. I. & Blaber, S. J. M. 1992 Plankton, epibenthos and
De Sylva, D. P. 1963 Systematics and life-history of the great fish communities. In Tropical Mangrove Ecosystems (Robertson,
barracuda Sphyraena barracuda. Studies in Tropical Oceanography A. I. & Alongi, D. M., eds). Coastal and Estuarine Studies
1, 179 pp. No. 41, 173–224.
Migration patterns and habitat preference of coral reef fishes 321
Robertson, D. R. & Warner, R. R. 1978 Sexual patterns in the Sponaugle, S. & Cowen, R. K. 1996 Larval supply and patterns
labroid fishes of the Western Caribbean, II: The parrotfishes of recruitment for two Caribbean reef fishes, Stegastus partitus
(Scaridae). In Smithsonian Contributions to Zoology (Robertson, and Acanthurus bahianus. Marine and Freshwater Research 47, 433–
D. R. & Warner, R. R., eds). Smithsonian Institute, Washington, 447.
DC, No. 255, 635 pp. Starck, W. A. & Schroeder, R. E. 1971 Investigations on the gray
Rooker, J. R. & Dennis, G. D. 1991 Diel, lunar and seasonal snapper Lutjanus griseus. Studies in Tropical Oceanography 10,
changes in a mangrove fish assemblage off southwestern Puerto 224 pp.
Rico. Bulletin of Marine Science 49, 684–698. Thayer, G. W., Colby, D. R. & Hettler, W. F. Jr 1987 Utilization of
Rutherford, E. S., Schmidt, T. W. & Tilman, J. T. 1989 Early life the red mangrove prop root habitat by fishes in south Florida.
history of spotted sea trout (Cynoscion nebulosus) and gray snapper Marine Ecology Progress Series 35, 25–38.
(Lutjanus griseus) in Florida Bay, Everglades National Park, Weinstein, M. P. & Heck, K. L. Jr 1979 Ichthyofauna of seagrass
Florida. Bulletin of Marine Science 44, 49–64. meadows along the Caribbean coast of Panama and in the gulf of
´
Sedberry, G. R. & Carter, J. 1993 The fish community of a shal- Mexico: composition, structure and community ecology. Marine
low tropical lagoon in Belize, Central America. Estuaries 16, Biology 50, 97–107.
198–215. Yanez-Arancibia, A., Lara-Dominguez, A. L., Rojas-Galaviz, J. L.,
´˜
Shulman, M. J. 1985 Recruitment of coral reef fishes: efects of Sanchez-Gil, P., Day, J. W. & Madden, C. J. 1988 Seasonal
´
distribution of predators and shelter. Ecology 66, 1056–1066. biomass and diversity of estuarine fishes coupled with tropical
Sokal, R. R. & Rohlf, F. J. 1995 Biometry, 3rd edition. WH habitat heterogeneity (southern Gulf of Mexico). Journal of Fish
Freeman and Company, New York, 887 pp. Biology 33 (Suppl A), 191–200.