The Importance of Mangroves, Mud and Sand Flats, and Seagrass Beds as Feeding Areas for Juvenile Fishes in Chwaka Bay, Zanzibar: Gut Content and Stable Isotope Analyses (Lugendo et al, 2006)
Journal of Fish Biology (2006) 69, 1639–1661
doi:10.1111/j.1095-8649.2006.01231.x, available online at http://www.blackwell-synergy.com
The importance of mangroves, mud and sand flats, and
seagrass beds as feeding areas for juvenile fishes in
Chwaka Bay, Zanzibar: gut content and stable
isotope analyses
B. R. LUGENDO*†, I. NAGELKERKEN†‡, G. VELDE†§
VAN DER
AND Y. D. MGAYA*
*Faculty of Aquatic Sciences and Technology, University of Dar es Salaam, P. O. Box
35064, Dar es Salaam, Tanzania, †Department of Animal Ecology and Ecophysiology,
Institute for Water and Wetland Research, Faculty of Science, Radboud University,
Toernooiveld 1, 6525 ED Nijmegen, The Netherlands and §National Museum of Natural
History, Naturalis, P.O. Box 9517, 2300 RA Leiden, The Netherlands
(Received 8 December 2005, Accepted 5 July 2006)
The relative importance of bay habitats, consisting of mangrove creeks and channel, seagrass beds,
and mud and sand flats, as feeding grounds for a number of fish species was studied in Chwaka
Bay, Zanzibar, Tanzania, using gut content analysis and stable isotope analysis of carbon and
nitrogen. Gut content analysis revealed that within fish species almost the same food items were
consumed regardless of the different habitats in which they were caught. Crustaceans (mainly
copepods, crabs and shrimps) were the preferred food for most zoobenthivores and omnivores,
while fishes and algae were the preferred food for piscivores and herbivores, respectively. The mean
d13C values of fishes and food items from the mangrove habitats were significantly depleted to
those from the seagrass habitats by 6Á9 and 9Á7% for fishes and food items, respectively, and to
those from the mud and sand flats by 3Á5 and 5Á8%, respectively. Fishes and food items from the
mud and sand flats were significantly depleted as compared to those of the seagrass habitats by 3Á4
and 3Á9%, for fishes and food, respectively. Similar to other studies done in different geographical
locations, the importance of mangrove and seagrass themselves as a primary source of carbon to
higher trophic levels is limited. The different bay habitats were all used as feeding grounds by
different fish species. Individuals of the species Gerres filamentosus, Gerres oyena, Lethrinus lentjan,
Lutjanus fulviflamma, Pelates quadrilineatus and Siganus sutor appeared to show a connectivity with
respect to feeding between different habitats by having d13C values which were in-between those of
food items from two neighbouring habitats. This connectivity could be a result of either daily tidal
migrations or recent ontogenetic migration. # 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles
Key words: feeding areas; habitat connectivity; juvenile fishes; mangroves; stable isotopes;
seagrass beds.
‡Author to whom correspondence should be addressed. Tel.: þ31 24 365 2471; fax: þ31 24 365 2409;
email: i.nagelkerken@science.ru.nl
1639
2006 The Authors
#
Journal compilation # 2006 The Fisheries Society of the British Isles
1640 B. R. LUGENDO ET AL.
INTRODUCTION
Mangrove and seagrass habitats are often characterized by high densities of
juvenile fishes and are therefore often referred to as nursery habitats (Robertson
& Duke, 1987; Little et al., 1988; Parrish, 1989), although little evidence has yet
been provided for this (Beck et al., 2001; Chittaro et al., 2004). Protection
against predation, a high food abundance and easy interception of planktonic
fish larvae due to the large areas of the habitats are among the assumptions
used in explaining the high abundances of juvenile reef fish species in these
habitats (Parrish, 1989; Robertson & Blaber, 1992). Few studies have, how-
ever, tested these hypotheses (Laegdsgaard & Johnson, 2001; Cocheret de la
`
Moriniere et al., 2004; Verweij et al., 2006) in contrast to numerous studies that
describe the fish assemblages of such habitats. The contradicting information
about the functioning of these habitats (Chong et al., 1990) creates a need to
investigate several regions independently. As pointed out by Hartill et al.
(2003), a better understanding is required of the resources used by different fish
species and life stages, and of how important different habitats are in maintain-
ing fish populations before management plans can be improved.
Mangrove and seagrass habitats are often interlinked through diurnal and
tidal fish migrations (Rooker & Dennis, 1991; Vance et al., 1996; Nagelkerken
et al., 2000; Dorenbosch et al., 2004). Little is known, however, of the degree to
which these habitats are used as feeding habitats (Nagelkerken & van der Velde,
2004). Conventional techniques such as gut content analysis may provide unreli-
able results with respect to the diet composition and the source of the food due
to the following reasons: 1) differences in digestion rates of ingested material, 2)
contents can be hard to identify, 3) not all contents are digested, 4) it provides
just a snapshot of the true diet and 5) it does not show from where the food
originates (MacDonald et al., 1982; Gearing, 1991; Polis & Strong, 1996).
Nonetheless, it proves to be the only means of establishing details of the types
and amounts of prey taken (Sydeman et al., 1997). Analysis of the stable iso-
topes of carbon and nitrogen can provide a clearer understanding of diets
because they reflect the actual assimilation of organic matter into consumer tis-
sue rather than merely its consumption, and provide an average of the diet over
periods of weeks to months (Gearing, 1991). The power of stable isotope anal-
ysis as a tool in the investigation of aquatic food web structures and dietary
patterns is based on the significant and consistent differences in isotopic com-
position of different types of primary producers due to different photosynthetic
pathways or different inorganic carbon sources (Bouillon et al., 2002a). The sta-
ble isotopic composition of an animal reflects that of its diet with up to 1Á0%
enrichment in 13C and an average of 3Á5% enrichment in 15N between a consumer
and its food source (DeNiro & Epstein, 1978; Fry & Sherr, 1984; Minagawa &
Wada, 1984) due to the discrimination against lighter isotopes during assimilatory
and excretory functions within consumers (Minagawa & Wada, 1984). The actual
degree of fractionation, however, varies as a function of taxonomy, food quality
and environmental factors (Vanderklift & Ponsard, 2003).
The aim of the present study was to establish the relative importance of dif-
ferent bay habitats, namely, mangroves, seagrass beds, and mud and sand flats,
as feeding areas for juveniles of a number of commercially important fish
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1641
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
species in Chwaka Bay, Zanzibar. The combination of gut analysis and stable
isotope analysis was expected to provide information on both the type and re-
lative amount of prey ingested and to reflect the sources of the food assimilated
by different fish species over periods of weeks up to months. This study endeav-
oured to answer the following questions: 1) Is there a significant difference in
stable isotopic signature (C and N) of fishes and food items in different bay
habitats? 2) In which habitats do fishes eat and what do they consume? 3) To
what degree does connectivity between habitats due to feeding by fishes exist?
MATERIALS AND METHODS
S T U D Y A R EA
The study was carried out in Chwaka Bay, a shallow bay located on the east coast of
Unguja Island, Zanzibar, Tanzania (Fig. 1). Chwaka Bay consists of a large intertidal
flat partly covered with mixed assemblages of algae and seagrass beds with an average
depth of 3Á2 m, an estimated area of 50 km2 at high spring tide and 20 km2 at low
spring tide, and a mean tidal range of 3Á2 m (Cederlof et al., 1995; Mahongo, 1997).
¨
Chwaka Bay is protected from the high-energy ocean on the east coast by a reef system
running along the coastline, as well as the Michamvi Peninsula (Fig. 1). On the land-
ward side, the bay is fringed by a dense mangrove forest of c. 3000 ha (Mohammed
et al., 2001). The mangrove forest has a number of tidal creeks fringed by prop roots
of the mangrove Rhizophora mucronata (Lamarck), with Mapopwe Creek (c. 2 m deep)
being the largest and the main water exchange route between the forest and the bay.
The mangrove creeks and the channel are intertidal in nature and none have any sig-
nificant freshwater input other than rain. The sampled habitats were: mangrove creeks,
mangrove channel, mud and sand flats, Chwaka seagrass beds (seagrass beds close to
the mangroves) and Marumbi seagrass beds (seagrass beds far from mangroves)
(Fig. 1). The sampled seagrass beds consisted of vast fields of Enhalus acoroides (L.)
39°24' 39°30'
N
4 8
0 km
6°6'
Tanzania Zanzibar
5
Marumbi
Mic
ham
Zanzibar
Chwaka
vi P
4
e
nin
Chwaka
3
sula
Bay
6°12' 2
1
FIG. 1. Map of Unguja Island (Zanzibar) showing the location of Chwaka Bay and the sampled habitats
(1, mangrove creeks; 2, mangrove channel; 3, mud and sand flats; 4, Chwaka seagrass beds; 5,
Marumbi seagrass beds). Grey areas in Chwaka Bay indicate mangrove forests.
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1642 B. R. LUGENDO ET AL.
˚
Royle interrupted by small patches of Thalassodendron ciliatum (Forsskal) den Hartog
and the calcareous algae Halimeda spp.
S A M P L IN G D ES I GN
Sample collections were carried out between November 2001 and October 2002. Fish
samples were collected using a seine, while macrofauna and macroflora samples were
collected by hand. Zooplankton samples were collected using a plankton net (80 mm
mesh). In the field, samples were put in a cool box and later frozen at À20° C pending
analysis. Fish species were selected in such a way that they represented commercially
important fish species found abundantly (see Table I) in more than one bay habitat,
and they included five feeding guilds: herbivores [Siganus sutor (Valenciennes)], insecti-
vores [Zenarchopterus dispar (Valenciennes)], omnivores [Monodactylus argenteus (L.)],
piscivores [Sphyraena barracuda (Walbaum)] and zoobenthivores [Gerres filamentosus
˚ `
Cuvier, Gerres oyena (Forsskal), Lethrinus lentjan (Lacepede), Lutjanus fulviflamma
˚
(Forsskal) and Pelates quadrilineatus (Bloch)]. Fish guild membership was assigned using
Smith & Heemstra (1991), Khalaf & Kochzius (2002) and Froese & Pauly (2004), which
were also used as a guide for the sampling of potential food items for each fish species.
Detailed information on the environmental variables and the fish community structure
(and their temporal variation) of Chwaka Bay can be found in other studies (Lugendo
et al., 2005, in press; B. R. Lugendo, I. Nagelkerken, N. S. Jiddawi, G. van der Velde
and Y. D. Mgaya, unpubl. data).
S T A B L E I S O T O PE A N A L Y SI S
Muscle tissues were removed from the fishes, while molluscs (gastropods and bi-
valves) and crustaceans (crabs and shrimps) were dissected from their exoskeleton
or shells prior to drying. The zooplankton samples were cleaned from detritus, sedi-
ments and other materials, under a dissecting microscope. Samples were dried at 70° C
for 48 h and ground to powder (homogeneous mixture). For samples rich in carbo-
nates such as detritus and whole individuals of small hermit crabs, sub-samples were
acid-washed and oven-dried. These sub-samples were used for stable carbon isotope
analysis only, while the remaining untreated sub-samples were used for stable nitrogen
isotope analysis since acid-washing interferes with stable nitrogen isotopes (Pinnegar &
Polunin, 1999). Samples were placed in ultra-pure tin capsules and combusted in a Carlo
ErbaÒ NA 1500 elemental analyser coupled on-line via a Finnigan Conflo III interface
with a ThermoFinnigan DeltaPlus mass spectrometer. Carbon and nitrogen isotope
ratios are expressed in the delta notation (d13C and d15N) relative to Vienna PDB
and atmospheric nitrogen. The potential food items and possible feeding habitat for
fishes were determined in view of the enrichment in isotope signatures of 1 and 3Á5%,
for carbon and nitrogen, respectively, between fishes and their potential food items
(DeNiro & Epstein, 1978; Minagawa & Wada, 1984). The term ‘macroinvertebrate’ is
used in the figures to denote zoobenthos and insects together, while the term ‘zoobenthos’
whenever used in the figures excludes the insects.
GUT CONTENT ANALYSIS
For fishes, fork length (LF) was measured to the nearest 0Á1 cm, and the entire gut
extracted and frozen pending analysis. The gut was then split, the gut contents placed
in a Petri dish under a dissecting microscope and food items were identified to the low-
est taxa possible. The percentage of the total stomach volume that each food category
comprised was determined using the point method (Hyslop, 1980) in which the food
items in each fish gut was allotted a number of points depending on its abundance
and size of an organism (i.e. one large organism counted as much as a large number
of small ones). The points and the percentages they represented were 5 (75–100%), 4
(50–75%), 3 (25–50%), 2 (5–25%) and 1 (up to 5%). All the points gained by each
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
TABLE I. Stable carbon and nitrogen isotope signatures (mean Æ S.E.) of different fork length (LF) classes of fish species in different bay
habitats. The overall mean d13C is also shown where more than one size class of fish species was present in a habitat. Relative abundance and
relative biomass for each species in each bay habitat are also given. Numbers in bold print show relative proportions of each species for the
whole bay. Different superscript lowercase letters and numbers represent statistical post hoc results and denote significantly different
(P < 0Á05) stable carbon isotope values of a fish species for similar LF classes and for overall d13C among different bay habitats
# 2006 The Authors
Overall Relative Relative
LF class
N
Species (cm) d13C d15N Species mean d13C abundance (%) biomass (%)
Gerres filamentosus Gerres filamentosus 7Á5 3Á6
Mangrove creeks 0–5 2 0Á3 8Á3 0Á1 Mangrove creeks 9Á6 5Á8
À21Á3 Æ Æ À21Á2 Æ 0Á4a
Mangrove creeks 5–10 3 0Á5a 8Á0 0Á1 Mangrove channel 24Á3 17Á0
À21Á1 Æ Æ À21Á6 Æ 0Á5a
Mangrove channel 5–10 5 0Á5a 8Á4 0Á1 Mud and sand flats 0Á8 1Á1
À21Á6 Æ Æ À19Á2 Æ 1Á0a
Mud and sand flats 5–10 4 1Á0a 8Á0 0Á2
À19Á2 Æ Æ
Gerres oyena Gerres oyena 22Á6 21Á2
Mangrove creeks 5–10 10 0Á3a 7Á5 0Á2 Mangrove creeks 7Á1 7Á1
À19Á4 Æ Æ
Mangrove channel 5–10 10 0Á7b 7Á4 0Á1 Mangrove channel 25Á6 35Á0
À17Á0 Æ Æ
Mud and sand flats 5–10 10 0Á3c 6Á6 0Á2 Mud and sand flats 62Á6 56Á2
À13Á8 Æ Æ
Chwaka seagrass beds 5–10 10 0Á7c 7Á5 0Á1 Chwaka seagrass beds 37Á6 38Á8
À12Á8 Æ Æ
Lethrinus lentjan Lethrinus lentjan 2Á7 1Á6
Mangrove channel 5–10 10 0Á3a 8Á0 0Á1 Mangrove channel 0Á3a 2Á4 1Á2
À21Á8 Æ Æ À21Á8 Æ
Mud and sand flats 5–10 9 0Á7b 6Á8 0Á2 Mud and sand flats 0Á7b 6Á8 3Á4
À19Á3 Æ Æ À19Á3 Æ
Chwaka seagrass beds 5–10 10 0Á2c 8Á0 0Á1 Chwaka seagrass beds 0Á2c 3Á9 3Á0
À12Á3 Æ Æ À12Á3 Æ
Marumbi seagrass beds 5–10 4 0Á6c 8Á3 0Á2 Marumbi seagrass beds 0Á4c 1Á7 1Á1
À12Á4 Æ Æ À12Á0 Æ
Marumbi seagrass beds 10–15 2 0Á1 8Á3 0Á0
À11Á6 Æ Æ
Lutjanus fulviflamma Lutjanus fulviflamma 2Á0 3Á3
Mangrove channel 5–10 4 0Á21 8Á5 0Á1 Mangrove creeks 0Á8 2Á0
À21Á0 Æ Æ À20Á1 Æ 0Á9a
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
Chwaka seagrass beds 5–10 4 0Á72 8Á0 0Á2 Mangrove channel 1Á4 2Á1
À15Á2 Æ Æ À21Á8 Æ 0Á1a
Mangrove creeks 10–15 3 0Á9ab 8Á6 0Á4 Mud and sand flats 3Á8 5Á6
À20Á1 Æ Æ À15Á2 Æ 0Á5b
Mangrove channel 10–15 2 0Á0a 9Á2 0Á0
À22Á6 Æ Æ
Mud and sand flats 10–15 5 0Á5c 7Á6 0Á2
À15Á2 Æ Æ
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1643
TABLE I. Continued
1644
Overall Relative Relative
LF class
N
Species (cm) d13C d15N Species mean d13C abundance (%) biomass (%)
Chwaka seagrass beds 10–15 5Á2 8Á2
2 À14Á2 Æ 0Á0bc 9Á0 Æ 0Á2 Chwaka seagrass beds À14Á5 Æ 0Á4b
Marumbi seagrass beds 10–15 0Á3 0Á4
Journal compilation
4 À11Á2 Æ 0Á4d 9Á0 Æ 0Á1 Marumbi seagrass beds À11Á2 Æ 0Á4c
#
Monodactylus argenteus Monodactylus argenteus 3Á1 1Á2
Mangrove creeks 0–5 5Á5 2Á8
11 À22Á1 Æ 0Á2a 8Á0 Æ 0Á1 Mangrove creeks
Mangrove channel 5–10 5Á5 3Á8
3 À24Á0 Æ 0Á6b 8Á4 Æ 0Á4 Mangrove channel
Pelates quadrilineatus Pelates quadrilineatus 2Á5 2Á4
Mud and sand flats 5–10 2Á9 1Á9
9 À17Á1 Æ 0Á1a 7Á2 Æ 0Á1 Mud and sand flats
Chwaka seagrass beds 5–10 12Á3 10Á8
10 À16Á2 Æ 0Á4b 7Á8 Æ 0Á2 Chwaka seagrass beds
Siganus sutor Siganus sutor 1Á6 3Á6
Mud and sand flats 5–10 7 0Á5a 1Á3 1Á2
À22Á8 Æ 5Á6 Æ 0Á3 Mud and sand flats À22Á8 Æ 0Á5a
Chwaka seagrass beds 5–10 4 0Á8a 2Á8 2Á0
À20Á7 Æ 7Á0 Æ 0Á2 Chwaka seagrass beds À19Á5 Æ 0Á7b
Marumbi seagrass beds 5–10 4 0Á7b 7Á4 11Á4
À15Á5 Æ 6Á7 Æ 0Á3 Marumbi seagrass beds À16Á1 Æ 0Á5c
Chwaka seagrass beds 10–15 1 6Á1
À15Á4
Marumbi seagrass beds 10–15 11 À16Á2 Æ 0Á6 6Á5 Æ 0Á1
Marumbi seagrass beds 15–20 11 À16Á5 Æ 0Á2 6Á3 Æ 0Á1
B. R. LUGENDO ET AL.
Sphyraena barracuda Sphyraena barracuda 0Á9 3Á8
Mangrove creeks 10–15 2 1Á01 1Á0a 0Á9 3Á1
À20Á6 Æ 9Á2 Æ 0Á0 Mangrove creeks À20Á6 Æ
Mangrove channel 10–15 5 0Á51 0Á5a 1Á1 4Á5
À19Á9 Æ 9Á8 Æ 0Á1 Mangrove channel À19Á9 Æ
Mud and sand flats 15–20 5 0Á2 0Á4b 1Á5 7Á2
À15Á7 Æ 8Á5 Æ 0Á2 Mud and flats À15Á9 Æ
Mud and sand flats 20–25 2 0Á5a 1Á8b 0Á7 5Á9
À16Á1 Æ 8Á2 Æ 0Á4 Chwaka seagrass beds À14Á6 Æ
Chwaka seagrass beds 20–25 2 1Á8a
À14Á6 Æ 9Á1 Æ 0Á6
Zenarchopterus dispar Zenarchopterus dispar 3Á9 4Á8
Mangrove creeks 10–15 8Á2 14Á6
9 À22Á8 Æ 0Á1a 8Á1 Æ 0Á1 Mangrove creeks
Mangrove channel 10–15 2Á9 4Á9
9 À22Á7 Æ 0Á1a 8Á2 Æ 0Á1 Mangrove channel
N, sample size.
2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
# 2006 The Authors
1645
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
food item were scaled down to percentages, to give percentage composition of each
food item in a diet of individual fish species examined.
STATISTICAL ANALYSIS
Each bay habitat was treated as a sample unit. First, data were pooled for each hab-
itat for fishes and for food items, respectively, in order to test for the overall differences
among habitats. Subsequently, each fish species was treated separately. The numbers of
individual fishes analysed for each particular species (i.e. sample size) equalled the num-
ber of replicates (N; Table I). Data were checked for homogeneity of variances using
a Levene’s test (Field, 2000). In case variances were homogeneous, a one-way ANOVA
or t-test was employed to test for differences in stable isotope signatures of carbon for
fishes and food items among different habitats. Since fish sample sizes were very differ-
ent (see Table I), a Hochberg’s GT2 was used as a post hoc test due to its greater sta-
tistical power in such kinds of data compared to other tests (Field, 2000). All data that
did not show homogeneous variances were log10-transformed, and a Levene’s test was
performed once again. Either Kruskal–Wallis test or Mann–Whitney U-test (depending
on the number of sample units involved) on the non-transformed data was used as a
non-parametric test equivalent when variances were not homogeneous, even after log10-
transformation. A Games–Howell post hoc test was used following the Kruskal–Wallis
tests because it is more powerful and specifically designed for lack of homogeneity of var-
iances (Field, 2000). A significance level of P < 0Á05 was used in all tests. All analyses
were performed using the programme SPSS 11.5 for Windows (Field, 2000).
RESULTS
GUT CONTENT ANALYSIS
Gut analysis indicated a food preference by different fish species, despite the
fact that they ingested a variety of food items (Table II). While some fish spe-
cies maintained a quite similar diet type regardless of the different habitats
from which they were caught (G. filamentosus: copepods; S. sutor: macroalgae;
S. barracuda: fishes; Z. dispar: insects), the diet of the other species (G. oyena,
L. lentjan, L. fulviflamma and M. argenteus) differed within species in different hab-
itats. The main food of G. oyena from the mangrove channel and from Chwaka
seagrass beds consisted mainly of copepods while fishes from mud and sand
flats fed mainly on detritus (Table II). Lethrinus lentjan fed mainly on ostracods
in the mangrove channel, on copepods on the mud and sand flats and on crus-
taceans and insects in the Chwaka seagrass beds. The diet of L. fulviflamma
consisted mainly of crustaceans in the mangroves, of copepods on the mud
and sand flats, of crabs and shrimps in Chwaka seagrass beds, and of crabs
and fishes in Marumbi seagrass beds. Monodactylus argenteus from the man-
grove creeks fed mainly on copepods while those from mangrove channel fed
mainly on algae (Table II).
M E A N d1 3 C S I G N A T U R E S F O R F I S H E S A N D F O O D I T E M S
A clear gradient in d13C could be discerned for fishes as well as food items
from the mangrove habitats located deep into the bay to the seagrass beds at
the mouth of the bay (Fig. 2). Fishes and food items from the mangrove hab-
itats were significantly depleted (Hochberg’s GT2, P < 0Á001) to those from the
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
TABLE II. Mean percentage composition of diet for different fish species and fork length (LF) classes in different bay habitats. Grey boxes
highlight all food items with a relative abundance of >19%.
1646
Unidentified Unidentified
Crustacean animal plant
LF class
Species/site (cm) N Copepod Crab Shrimp Ostracod Parts Fishes Detritus Gastropod Nematode Insect Algae Seagrass Sediment material material Other
Journal compilation
Gerres filamentosus (ZB)
#
Mangrove 0–5, 5–10 3 95Á8 1Á0 3Á2
creeks
Mangrove 0–5, 5–10 45 71Á2 1Á9 0Á3 11Á4 0Á1 2Á3 9Á9 2Á8
channel
Gerres oyena (ZB)
Mangrove 5–10 15 42Á9 2Á0 15Á2 7Á1 0Á2 9Á8 12Á5 6Á2 4Á2
channel
Mud and 0–5, 5–10, 16 21Á2 2Á1 53Á3 8Á4 0Á2 1Á9 8Á7 4Á2
sand flats 10–15
Chwaka 10–15 11 39Á2 2Á7 26Á1 3Á4 4Á8 11Á2 12Á6
seagrass beds
Lethrinus lentjan (ZB)
Mangrove 5–10 7 12Á9 53Á6 14Á3 2Á1 17Á1
channel
Mud and 5–10 12 70Á8 3Á1 8Á3 0Á3 0Á3 17Á0 0Á3
sand flats
Chwaka 5–10, 10–15 3 33Á3 33Á3 33Á3
B. R. LUGENDO ET AL.
seagrass beds
Lutjanus fulviflamma (ZB)
Mangrove 5–10 6 19Á2 19Á2 27Á1 16Á7 16Á7 1Á2
channel
Mud and 5–10, 10–15 12 39Á6 8Á3 20Á8 0Á3 8Á3 8Á3 14Á3
sand flats
Chwaka 5–10, 10–15 30 40Á4 22Á7 10Á4 0Á1 4Á2 20Á1 2Á1
seagrass beds
Marumbi 10–15 3 45Á8 20Á8 33Á3
seagrass beds
Monodactylus argenteus (O)
Mangrove 0–5, 5–10 18 46Á0 50* 0Á8 3Á0 0Á2
creeks
2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
# 2006 The Authors
TABLE II. Continued
Unidentified Unidentified
# 2006 The Authors
Crustacean animal plant
LF class
Species/site (cm) N Copepod Crab Shrimp Ostracod Parts Fishes Detritus Gastropod Nematode Insect Algae Seagrass Sediment material material Other
Mangrove 5–10 10 6Á3 17Á8 10Á0 65Á0 0Á9
channel
Pelates quadrilineatus (ZB)
Mud and 5–10, 10–15 5 32Á5 27Á5 20Á0 20Á0
sand flats
Zenarchopterus dispar (ZB)
Mangrove 10–15, 15–20 9 0Á3 11Á4 9Á7 69Á4 4Á2 4Á2 0Á7
creeks
Mangrove 15–20 5 20Á0 20Á0 40Á0 20Á0
channel
Siganus sutor (H)
Mangrove 5–10 5 92Á5 7Á5
channel
Mud and 5–10, 10–15 5 4Á4 72Á5 7Á5 15Á6
sand flats
Marumbi 5–10, 10–15, 23 7Á1 79Á4 2Á3 11Á2
seagrass beds 15–20
Sphyraena barracuda (P)
Mangrove 10–15 5 100Á0
channel
Mud and 10–15, 15–20, 7 16Á7 62Á5 1Á0 18Á8 1Á0
sand flats 20–25
Chwaka 10–20, 20–30 4 21Á9 1Á0 71Á9 1Á0 1Á5 2Á7
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
seagrass beds
H, herbivore; O, omnivore; P, piscivore; ZB, zoobenthivore; N, number of fish analysed; *, shrimp larvae.
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1647
1648 B. R. LUGENDO ET AL.
seagrass habitats by on average 6Á9 and 9Á7% for fishes and food items, respec-
tively, and from the mud and sand flats by on average 3Á5 and 5Á8%, respec-
tively. Food items from the mud and sand flats were significantly depleted
(Hochberg’s GT2, P < 0Á01) as compared to those of the seagrass habitats
by on average 3Á9%. Fishes from the mud and sand flats were depleted by
an average of 3Á4%, but this difference was not significant (Hochberg’s GT2,
P > 0Á05). There were no significant differences (Hochberg’s GT2, P > 0Á05)
in d13C between the two mangrove habitats (average difference of 0Á2 and
1Á5%, for fishes and food, respectively), and between the two seagrass habitats
(average difference of 1Á9 and 0Á2%, for fishes and food, respectively).
TROPHIC LEVELS OF FISHES AND FOOD ITEMS
The food web in the bay showed various trophic levels. Detritus and plant
material were generally more depleted in d15N as compared to zooplankton
and macroinvertebrates (zoobenthos þ insects) found within the same habitat,
while fishes were the most enriched in d15N (Fig. 3). Also, clear gradients in
both d13C and d15N could be observed for different feeding guilds of fishes
and for different habitats [Fig. 3(b)]. Three trophic levels could be discerned
for the fishes, with increasing values of d15N from herbivores to omnivores
and zoobenthivores to piscivores. For each feeding guild, d13C increased along
the spatial gradient from mangroves in the bay to seagrass beds at the mouth
of the bay [Fig. 3(b)].
S T A B L E I S O T O PI C S I G N A T U R E S O F F I SH S PE C I E S
Individual fish species from the mangrove habitats were generally more
depleted in d13C compared to those of the same species from either mud and
9
8
7
6
N
15
5
4
3
2
–26 –25 –24 –23 –22 –21 –20 –19 –18 –17 –16 –15 –14 –13 –12
13
C
FIG. 2. Pooled mean Æ S.E. stable carbon and nitrogen isotope values of fishes (), u, n, s, *) and food
items (r, n, m, d, Â) in different bay habitats. (), r, mangrove creeks; u, n, mangrove channel;
n, m, mud and sand flats; s, d, Chwaka seagrass beds; *, Â, Marumbi seagrass beds).
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1649
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
(a)
7
6
Zoobenthos
5
Zoobenthos
4 Macro-
Mangroves + Zooplankton Zooplankton
invertebrates
seagrasses
Zoobenthos
Zooplankton
Detritus
Detritus
Algae +
3 Zooplankton Macro- seagrasses
Zooplankton
Algae +
invertebrates
seagrasses
Algae +
2 seagrasses
Detritus
Detritus
Algae +
1 seagrasses
0
–28 –26 –24 –22 –20 –18 –16 –14 –12
N
15
13
C
(b)
11
Mangrove channel
10
Mangrove Chwaka
creeks seagrass beds Marumbi
Mangrove
9 seagrass beds
channel
Mangrove creeks
Mud and sand flats
8 Mangrove
Mangrove
Mud and sand flats
channel Chwaka
creeks
seagrass beds
7
Chwaka
seagrass beds Marumbi
6 seagrass beds
Mud/sand flats
5
–25 –23 –21 –19 –17 –15 –13 –11
13
C
FIG. 3. Mean d13C and d15N values of different (a) trophic groups of food items in different bay habitats.
(), mangrove creeks; u, mangrove channel; n, mud and sand flats; s, Chwaka seagrass beds; *,
Marumbi seagrass beds), and (b) feeding guilds of fish in different bay habitats (n, piscivores; u,
omnivores; ), zoobenthivores; s, herbivores).
sand flats or the seagrass habitats (Table I). The d13C depletion of individual
species was generally in the order: mangrove habitats < mud and sand flats
< seagrass habitats. The highest enrichment in d13C between two neighbouring
habitats was observed for individuals of the same LF class (10–15 cm) of L.
fulviflamma (mangrove channel and mud and sand flats: 7Á4%). With regard
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1650 B. R. LUGENDO ET AL.
to d15N, the herbivore S. sutor was the most depleted and the piscivore S. bar-
racuda the most enriched fish species, with a range of 2Á6–2Á8% (considering
the overall mean for LF classes) between the two species when occurring in the
same habitat. Considering individuals of the same species and similar size classes
in different bay habitats, d13C of five species, namely, G. oyena (5–10 cm), L.
lentjan (5–10 cm), L. fulviflamma (10–15 cm), P. quadrilineatus (5–10 cm) and
S. sutor (5–10 cm), differed significantly between habitats (one-way ANOVA,
d.f. ¼ 3,36, P < 0Á001 for G. oyena, Kruskal–Wallis, d.f. ¼ 3, P < 0Á001 for
L. lentjan, Kruskal–Wallis, d.f. ¼ 4, P < 0Á01 for L. fulviflamma, t-test, d.f. ¼ 1,
P < 0Á05 for P. quadrilineatus and one-way ANOVA, d.f. ¼ 2,12, P < 0Á001
for S. sutor), while those of G. filamentosus (5–10 cm) and Z. dispar (10–15 cm)
did not differ significantly between different bay habitats (Kruskal–Wallis,
d.f. ¼ 2, P > 0Á05 for G. filamentosus, Mann–Whitney U-test, d.f. ¼ 1, P >
0Á05 for Z. dispar). Similar results were obtained when different LF classes of indi-
vidual species were pooled within each habitat in which case also M. argenteus
differed significantly between the two mangrove habitats (t-test, d.f. ¼ 1, P <
0Á01). The post hoc results are presented in Table I.
ANALYSIS OF POTENTIAL FOOD AND FEEDING
H A B I T A T S O F D I F F E R E N T F I S H S P E C I ES
The herbivore S. sutor ingested mainly macroalgae (Table II). The d13C and
15
d N values of the average diet of this species were generally quite similar to
those of macroalgae, but very distinct from those of seagrasses or mangrove
leaves [Fig. 4(a)]. Siganus sutor from the mud and sand flats showed stable iso-
tope signatures indicating various types of macroalgae from the mangrove chan-
nel as a potential food source, while fish from Chwaka seagrass beds showed
values indicating green and brown algae from mud and sand flats and the cal-
careous green algae (Halimeda sp.) from Chwaka seagrass beds as a potential
food source. Siganus sutor from Marumbi seagrass beds showed an intermediate
value for its average diet that lay in-between those of green algae from the mud
and sand flats, Halimeda sp. from Chwaka seagrass beds, and calcareous green
algae (Udotea sp.) and red algae from Marumbi seagrass beds.
The gut content of the insectivore Z. dispar showed that insects formed a major
part of its diet (Table II), while the stable isotope values from both mangrove
habitats suggested a mixed diet of crabs (Sesarma sp. and Portunidae), shrimps
and insects from the mangroves [Fig. 4(b)].
FIG. 4. Mean Æ S.E. d13C and d15N values of fish species (a) Siganus sutor, (b) Zenarchopterus dispar, (c)
Monodactylus argenteus, (d) Sphyraena barracuda, (e) Gerres filamentosus, (f) Gerres oyena, (g)
Lethrinus lentjan, (h) Lutjanus fulviflamma and (i) Pelates quadrilineatus (large symbols) and potential
food items (small symbols) in different bay habitats. (), mangrove creeks; u, mangrove channel; n,
mud and sand flats; s, Chwaka seagrass beds; *, Marumbi seagrass beds). The arrow heads indicate
the predicted average d13C and d15N values (based on the 1 and 3Á5% enrichment, respectively, in
d13C and d15N between an animal and its food source) of the diet of fishes. The dashed lines combine
potential food sources within a habitat. Prey species are depicted on lowest taxonomic level for each
habitat in which the fish species was found; for the remainder of the habitats the prey species are
pooled to higher taxonomic levels (e.g. macroinvertebrates, zoobenthos and seagrasses).
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1651
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
(a)
8
6
Red algae
Red algae
Green algae
Mangrove
Thalassia
Green
Green
leaves
4 hemprichii
algae
algae Enhalus
Brown
Brown Udotea sp. acoroides
algae
Green algae
Red algae Thalassodendron
Red algae algae
ciliatum
Thalassodendron
Halimeda sp. Thalassia
ciliatum
2 Eucheuma sp. hemprichii
Halimeda Thalassia
Jania sp.
sp. hemprichii
Halimeda
Halodule sp. Enhalus acoroides Enhalus
wrightii
Cymodocea sp. acoroides
0
Brown
algae
Halodule
wrightii
–2
–30 –28 –26 –24 –22 –20 –18 –16 –14 –12 –10 –8 –6
(b)
9
7 Insects
Insects
Sesarma sp.
5 Animals
Shrimps
N
Portunidae Animals
15
Terebralia sp.
Red algae
Zooplankton
Detritus Brown algae Uca spp. Animals
3 Zooplankton
Algae +
seagrasses Algae +
Green algae
Detritus seagrasses
Cyanobacteria
1
Algae +
Seagrasses
Halodule wrightii seagrasses
–1
–30 –28 –26 –24 –22 –20 –18 –16 –14 –12 –10
(c)
10
8
Insects
Sesarma Insects Zoobenthos
6 sp.
Shrimps Zooplankton
Zoobenthos
Portunidae
4 Terebralia sp. Zooplankton
Red algae
Zooplankton Brown Zoobenthos
Detritus
algae
Detritus Algae +
Green algae Algae +
seagrasses
Zooplankton seagrasses
Zooplankton
Uca
2 spp.
Algae +
Halimeda sp.
seagrasses
Detritus Detritus
Thalassia
hemprichii
Halodule wrightii
0
Halodule wrightii
–2
–30 –28 –26 –24 –22 –20 –18 –16 –14 –12 –10
13
C
FIG. 4. Continued
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1652 B. R. LUGENDO ET AL.
11 (d)
10
9
Omnivores Carnivores
Carnivores
8 Carnivores
Omnivores Carnivores
Carnivores
7 Herbivores
?
Herbivores
6 ?
Herbivores
? ? Zoobenthos
?
?
?
5
Zoobenthos
4 Zoobenthos
Zoobenthos
Zoobenthos
3
–25 –23 –21 –19 –17 –15 –13 –11
9 (e)
8
7
Benthic
invertebrates
Bivalve sp.1
6 Sesarma sp.
N
Gastropods
Shrimps
5 Polychaetes
15
Benthic invertebrates
Zooplankton Terebralia sp.
4 Portunidae
Shrimps
Zooplankton
Terebralia sp.
Zooplankton
3 Zooplankton
Zooplankton Bivalve sp.2
Uca spp.
Hermit crabs
2 Dotilla
Crab sp. fenestrata
1
–28 –26 –24 –22 –20 –18 –16 –14 –12 –10
9 (f)
8
7 Isopods
Cypraea sp.
Brittle stars
6 Polychaetes
Sesarma sp. Bivalve sp.1
Thalamita sp.
Gastropods
Shrimps Zooplankton
5 Polychaetes
Zoobenthos
Amphipods
Portunidae
Shrimps
4
Hermit crabs
Zooplankton
Zooplankton Terebralia sp.
Terebralia sp. Dotilla
3 fenestrata
Zooplankton Zooplankton
Uca spp. Bivalve sp.2
Hermit crabs
2
Crabs
Dotilla
1 fenestrata
0
–26 –24 –22 –20 –18 –16 –14 –12 –10
13
C
FIG. 4. Continued
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1653
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
10 (g)
8 Littoraria sp.
Brittle stars
Isopods
Cypraea sp.
Polychaetes
Brittle stars
6 Sesarma sp. Bivalve sp.1
Thalamita sp.
Gastropods Polychaetes
Shrimps
Zooplankton Pinna sp.
Amphipods
Polychaetes
Shrimps
4 Hermit crabs
Portunidae Terebralia sp. Zooplankton Dotilla fenestrata
Terebralia sp. Hermit crabs
Zooplankton Zooplankton
Zooplankton Uca spp. Gastropods
Bivalve sp.2
Hermit
2 crabs
Crabs
Dotilla
fenestrata
0
–28 –26 –24 –22 –20 –18 –16 –14 –12 –10 –8
10 (h)
8 Littoraria sp.
Isopods Cypraea sp.
Brittle stars
Polychaetes
Brittle stars
6 Sesarma sp.
Thalamita sp.
Polychaetes
N
Pinna sp. Gastropods
15
Zooplankton
Polychaetes Amphipods
Shrimps
Portunidae Shrimps
Hermit crabs
4 Bivalves
Zooplankton Dotilla
fenestrata
Zooplankton Hermit
Terebralia sp. Terebralia sp.
crabs
Zooplankton
Zooplankton Zooplankton
Dotilla
Uca spp.
2 Hermit crabs fenestrata
Crabs
Dotilla
fenestrata
0
–28 –26 –24 –22 –20 –18 –16 –14 –12 –10 –8
(i)
9
8
7 Brittle stars
Brittle stars
6
Zoobenthos
5 Pinna sp.
Zooplankton
Gastropod sp.
4 Zooplankton Zoobenthos
Zooplankton Zoobenthos
3 Zoobenthos
Zooplankton
Zooplankton
Zoobenthos
Hermit crabs
2
1
0
–28 –26 –24 –22 –20 –18 –16 –14 –12 –10
13
C
FIG. 4. Continued
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1654 B. R. LUGENDO ET AL.
The omnivore M. argenteus ingested zooplankton, algae and some detritus
(Table II). The ingestion of zooplankton and detritus is supported by the
d13C for fish from the mangrove channel, although the enrichment in d15N
was larger than the usual 3Á5% [Fig. 4(c)]. For fish from the mangrove creeks,
the d13C signature suggests the diet to consist of a mixture of decapods (Sesarma
sp., Portunidae and shrimps) from the mangrove creeks and zooplankton and
detritus from the mangrove channel, but without an indication of dependence
on algae as a food source [Fig. 4(c)].
Although the gut content analysis shows that fishes formed major part of the
diet of the piscivore S. barracuda, the stable isotope signatures of the average
diet of S. barracuda was not close enough to those of the selected fish species
of this study to depend solely on these species as a food source. Sphyraena
barracuda from the mangrove habitats had an isotope signature of its average
diet that was closest to that of herbivorous fishes from the mud and sand flats
and Chwaka seagrass beds, while for S. barracuda from the mud and sand flats
and Chwaka seagrass beds this was the case for herbivorous fish from the Mar-
umbi seagrass beds, with a possibility of feeding partly on macrofauna too
[Fig. 4(d)].
In conformity with the gut content analysis where crustaceans (mainly cope-
pods, crabs and shrimps) formed a major part of the diet of most zoobenthi-
vores (Table II), G. filamentosus from the mangrove habitats had stable
isotope signatures for its average diet which lay in-between those of crustaceans
(Sesarma sp., Portunidae and shrimps) from the mangrove creeks, while fish
from mud and sand flats had isotope signatures for their average diet which
lay in-between values for shrimps from the mangrove creeks and zooplankton
from mud and sand flats [Fig. 4(e)]. Gerres oyena from the mangrove creeks
showed an isotope signature of its average diet close to the signatures of
shrimps from the mangrove creeks, gastropods (Terebralia sp.) from the man-
grove channel and zooplankton from the mud and sand flats, while G. oyena
from mangrove channel had signatures closest to zooplankton from the mud
and sand flats [Fig. 4(f)]. Gerres oyena from the mud and sand flats showed
an average diet signature close to that of bivalves, gastropod (Terebralia sp.)
and shrimps from mud and sand flats and zooplankton from the seagrass beds.
Gerres oyena from Chwaka seagrass beds showed a signature of its average diet
close to that of shrimps and gastropods (Terebralia sp.) from the mud and sand
flats, hermit crabs and amphipods from the Chwaka seagrass beds, and zoo-
benthos from the Marumbi seagrass beds [Fig. 4(f)]. Lethrinus lentjan from
the mangrove channel showed an isotope signature of its average diet that
was intermediate between crabs and shrimps of the mangrove creeks, while
for the mud and sand flats the signatures suggested a possible mix of shrimps,
crabs (Uca spp.) and gastropod (Terebralia sp.) from the mangroves and hermit
crabs and zooplankton from the mud and sand flats as a food source [Fig. 4(g)].
Lethrinus lentjan from the seagrass habitats showed an average stable iso-
tope signature for its diet that was close to that of the zoobenthos from the
seagrass habitats. The isotope signature of the average diet of L. fulviflamma
from the mangrove habitats showed proximity to isotope signatures of crabs
and shrimps from the mangrove habitats, while that of fish from the mud
and sand flats and Chwaka seagrass beds suggested an intermediate isotope
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1655
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
signature of polychaetes, shrimps and zooplankton from the mud and sand flats
[Fig. 4(h)]. Lutjanus fulviflamma from the Marumbi seagrass bed showed a stable
isotope signature of its diet in close proximity to zoobenthos from the seagrass
beds. Pelates quadrilineatus from both the mud and sand flats and Chwaka sea-
grass beds showed isotope signatures for their average diet close to zooplankton
from mud and sand flats and the two seagrass beds [Fig. 4(i)].
C O N N E C T I V I T Y B ET W E E N H A B I T A T S
The isotopic signatures of the fish species in relation to that of the possible
food items suggest four possibilities of feeding connectivity between adjacent
bay habitats (Fig. 4): 1) connectivity between the two mangrove habitats for
G. filamentosus, L. lentjan, L. fulviflamma, M. argenteus and Z. dispar, 2) connec-
tivity between mangrove habitats and mud and sand flats for G. filamentosus,
L. lentjan and S. sutor, 3) connectivity between mud and sand flats and seagrass
habitats for G. oyena, L. fulviflamma, P. quadrilineatus and S. sutor, and 4) con-
nectivity between the two seagrass habitats for L. fulviflamma and L. lentjan.
DISCUSSION
Both gut content and stable carbon isotope analyses showed evidence that
the studied fish species generally relied as a food source on algae (herbivores)
and macroinvertebrates (omnivores and zoobenthivores), with crustaceans
(crabs, shrimps and copepods) playing a major role. The different d13C or
d15N values of the piscivore S. barracuda from those of herbivorous fishes
indicate a possible dependence for juveniles (10–25 cm) of this species on other
animals than fishes alone. Copepods were found to some degree in the guts of
the juveniles. In a study in Gazi Bay (Kenya), de Troch et al. (1998) identified
other animals like gammaridean amphipods, mysids, crabs and shrimps in the
stomachs of piscivorous fishes (including S. barracuda), an observation that in-
dicates that at juvenile stages S. barracuda is not solely piscivorous.
Although the stable isotope signatures showed evidence for food dependence
of the studied fish species on mangrove and seagrass habitats, the direct con-
sumption of either mangrove or seagrass leaves seemed to be absent or very
low. The mean d13C of mangrove leaves of À28Á1% is similar to the overall
values for mangrove leaves recorded in the Caribbean, India, Malaysia and
in Kenya (Rao et al., 1994; Chong et al., 2001; Bouillon et al., 2002a; Cocheret
`
de la Moriniere et al., 2003). Similar to what was observed by Sheaves & Molony
(2000), Bouillon et al. (2002b) and Kieckbusch et al. (2004), however, this
value is much more depleted as compared to either fish species (Sheaves &
Molony, 2000; Kieckbusch et al., 2004; this study) or to most of the macroin-
vertebrates (Hsieh et al., 2002; Bouillon et al., 2002b; Guest & Connolly, 2004;
Kieckbusch et al., 2004; Abed-Navandi & Dworschak, 2005) so as to function
as a (direct and significant) source of carbon for these fauna. The most
depleted fish species in this study was M. argenteus with a mean d13C of
À24Á0%, which is far more enriched as compared to mangrove leaves. Simi-
larly, Guest & Connolly (2004) in Moreton Bay (Australia), Macia (2004) in
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1656 B. R. LUGENDO ET AL.
Inhaca Island (Mozambique) and Abed-Navandi & Dworschak (2005) on the
Belize Barrier Reef (Caribbean Sea) observed that the d13C of most crabs
and shrimps from the mangrove habitats was close to that of microphytoben-
thos and distinct from that of mangrove leaves.
Seagrasses (with exception of Halodule wrightii in the mangrove creeks and
channel with a mean d13C of À26Á3 and À20Á2%, respectively) were too far en-
riched in d13C (À15Á5 to À8Á2%) as compared to the herbivore S. sutor. This
suggests that seagrasses did not contribute to the diet of this herbivorous fish
species. The low contribution of seagrasses and the high contribution of algae
to the food web that was observed by Moncreiff & Sullivan (2001) in the Gulf
of Mexico and by Kieckbusch et al. (2004) in Biscayne Bay is another example
that seagrass plays a minor role in the food web and that algae are the primary
source of organic matter for higher trophic levels. Mangroves and seagrasses
do not appear to be direct sources of carbon in the diets of the fish species
studied; they probably serve as refugia as well as a substratum for a variety
of primary producers and consumers that are important in the food webs of
these habitats (Kieckbusch et al., 2004). Presence of food in addition to struc-
tural complexity has been reported to account for the strong association of
large numbers of juvenile fishes within mangrove forests (Laegdsgaard & Johnson,
2001). In addition, seagrass beds have also been reported to harbour a high
abundance of small invertebrates that are an important food of many juvenile
fish species (Nakamura & Sano, 2005).
Using stable carbon isotope analysis different habitats were distinguished,
which functioned as a source of carbon. Fish species from the mangroves were
more depleted in d13C as compared to individuals of the same species caught
from either the mud and sand flats or seagrass habitats. Similarly, fish species
from the mud and sand flats were more depleted relative to individuals of the
same species occurring in seagrass beds. The d13C of food also showed this
trend. This is in agreement with other studies showing that the importance
of mangrove-derived carbon (if any) is limited to the surroundings of the man-
grove habitats, and decreases when moving away from the mangroves (Rodelli
et al., 1984; Newell et al., 1995; Dehairs et al., 2000; Chong et al., 2001; Guest
& Connolly, 2004). In agreement with Dehairs et al. (2000), this observation
calls for critical evaluation on the assumption that mangrove ecosystem repre-
sent a source of organic nutrients for the coastal ecosystems. Like in other
studies from around the world, the present study shows significant feeding of
fishes (and macrobenthos) in the mangroves (Rodelli et al., 1984; Marguillier
`
et al., 1997; Sheaves & Molony, 2000; Chong et al., 2001; Cocheret de la Moriniere
et al., 2003; Guest & Connolly, 2004; Nagelkerken & van der Velde, 2004;
Abed-Navandi & Dworschak, 2005).
The overlap in stable carbon isotopes of some fish species in different bay
habitats suggests connectivity between these habitats, with the possibility that
fishes used more than one habitat as a feeding ground. Some fish species (G.
filamentosus, L. lentjan and S. sutor) from the mud and sand flats showed a pos-
sible connection to the mangrove habitats as feeding habitats. Likewise, some
fish species (G. oyena, L. fulviflamma, P. quadrilineatus and S. sutor) from Chwaka
seagrass beds showed some evidence of using mud and sand flats as feeding
habitats. An explanation for this observation could firstly be recent ontogenetic
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1657
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
`
migration (Cocheret de la Moriniere et al., 2003). The fishes could have
migrated from one habitat to another habitat recently, as a result of which they
still show part of the signature of their previously used habitat. It could take
several weeks to months to acquire the signature of the food from the new hab-
itat (Gearing, 1991; Hobson, 1999; Nagelkerken & van der Velde, 2004).
Since the fish samples were collected during low tide, a second possibility is
that fishes migrated with the tides (with a spring tidal difference of 2 m) from
the mangroves to the mud and sand flats and from the mud and sand flats to
Chwaka seagrass beds. Migration in relation to feeding (Reis & Dean, 1981),
preference of particular salinities (Quinn & Kojis, 1987) and avoidance of being
stranded during low tide in areas that fall dry (van der Veer & Bergman, 1986)
has been suggested to be among the reasons that can trigger tidal migrations.
Since the d13C signature showed intermediate values between habitats, this
could suggest that they fed at low tide as well as high tide, in two different hab-
itats. Tidal migration between bay habitats in Chwaka Bay by Lutjanidae has
been shown by Dorenbosch et al. (2004), and other species possibly follow the
same pattern of behaviour.
Distance to be covered during (tidal) migration, however, seems important in
terms of energy budget especially when juvenile fishes (<20 cm length) are con-
sidered, in which case long-distance migration costs may exceed energy intake
(Nøttestad et al., 1999). This may also be the case in the present study (in
which the majority of the fishes were 5–10 cm LF) where there appears to be
substantial connectivity for fish species between neighbouring habitats, but
not between habitats that were located far away from one another, such as
Marumbi seagrass beds located 8 and 6 km away from the mangrove and
mud and sand flat habitats, respectively.
The significant difference observed for some species in stable carbon isotopes
in individuals of the same species and similar size classes between bay habitats
suggests two situations: 1) the individuals of each habitat belong to different
assemblages, each depending completely (in terms of nutrition) on different
bay habitats, and 2) the different bay habitats all have the potential of provid-
ing sufficient food sources to the fish assemblage found therein. The differences
in fish densities of particular species and size class in different bay habitats as
observed by Lugendo et al. (2005), however, suggests that other factors than
food alone control the distribution of juvenile fishes. As observed from other
studies, structural complexity and shade in relation to predation risk are
among the important factors in determining distribution of juvenile fishes
`
(Laegdsgaard & Johnson, 2001; Cocheret de la Moriniere et al., 2004; Verweij
et al., 2006).
In conclusion, this study revealed that significant differences in stable isotope
signatures (C and N) exist in food and fishes from different bay habitats in
Chwaka Bay, which could be used to delineate feeding habitats of fishes. Fishes
appear to forage in all studied bay habitats. Seagrasses and mangroves do not
appear to be direct sources of carbon in the diets of studied fish species; rather,
they probably serve as refuge as well as a substratum for a variety of primary
producers and consumers that are important in the food webs of these habitats.
Some fish species of similar feeding guilds showed some degree of segregation
by feeding on different food resources. Zoobenthivores, however, showed an
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1658 B. R. LUGENDO ET AL.
overlap in diet and mainly fed on copepods, shrimps and crabs. There appears
to exist a connectivity for some fish species between different bay habitats with
respect to feeding (between the mud and sand flats and the mangroves, and
between the seagrass beds and the mud and sand flats), which could be a result
of either ontogenetic or tidal migration.
Gratitude is expressed to M. A. Manzi, S. J. Simgeni and M. A. Makame for assis-
tance with the fieldwork. The administration and staff of the Institute of Marine Sciences
in Zanzibar provided logistical support and research facilities. This study was financially
supported by NUFFIC through the ENVIRONS-MHO Project implemented by the Fac-
ulty of Science, University of Dar es Salaam, Tanzania, Schure-Beijerinck-Popping Foun-
dation and Quo Vadis Fonds (Radboud University), The Netherlands. I. N. was
supported by a Vidi grant from the Netherlands Organisation for Scientific Research
(NWO). This is Centre for Wetland Ecology publication no. 415.
References
Abed-Navandi, D. & Dworschak, P. C. (2005). Food sources of tropical thalassidean
shrimps: a stable-isotope study. Marine Ecology Progress Series 291, 159–168.
Beck, M. W., Heck, K. L. Jr, Able, K. W., Childers, D. L., Eggleston, D. B., Gillanders,
B. M., Halpern, B., Hays, C. G., Hoshino, K., Minello, T. J., Orth, R. J., Sheridan,
P. F. & Weinstein, M. P. (2001). The identification, conservation, and management
of estuarine and marine nurseries for fish and invertebrates. BioScience 51,
633–641.
Bouillon, S., Koedam, N. & Dehairs, F. (2002a). Primary producers sustaining macro-
invertebrate communities in intertidal mangrove forests. Oecologia 130, 441–448.
Bouillon, S., Raman, A. V., Dauby, P. & Dehairs, F. (2002b). Carbon and nitrogen stable
isotope ratios of subtidal benthic invertebrates in an estuarine mangrove ecosystem
(Andhra Pradesh, India). Estuarine Coastal and Shelf Science 54, 901–913.
Cederlof, U., Rydberg, L., Mgendi, M. & Mwaipopo, O. (1995). Tidal exchange in
¨
a warm tropical lagoon: Chwaka Bay, Zanzibar. Ambio 24, 458–464.
Chittaro, P. M., Fryer, B. J. & Sale, R. (2004). Discrimination of French grunts
(Haemulon flavolineatum Desmarest, 1823) from mangrove and coral reef habitats
using otolith microchemistry. Journal of Experimental Marine Biology and Ecology
308, 168–183.
Chong, V. C., Sasekumar, A., Leh, M. U. C. & Cruz, R. D. (1990). The fish and prawn
communities of a Malaysian coastal mangrove ecosystem, with comparisons to
adjacent mud/sand flats and inshore waters. Estuarine Coastal and Shelf Science
31, 703–722.
Chong, V. C., Low, C. B. & Ichikawa, T. (2001). Contribution of mangrove detritus to
juvenile prawn nutrition: a dual stable isotope study in a Malaysian mangrove
forest. Marine Biology 138, 77–86.
`
Cocheret de la Moriniere, E., Pollux, B. J. A., Nagelkerken, I., Hemminga, M. A.,
Huiskes, A. H. L. & van der Velde, G. (2003). Ontogenetic dietary changes of coral
reef fishes in the mangrove-seagrass-reef continuum: stable isotope and gut-content
analysis. Marine Ecology Progress Series 246, 279–289.
`
Cocheret de la Moriniere, E., Nagelkerken, I., van der Meij, H. & van der Velde, G.
(2004). What attracts juvenile coral reef fish to mangroves: habitat complexity or
shade? Marine Biology 144, 139–145.
Dehairs, F., Rao, R. G., Chandra Mohan, P., Raman, A. V., Marguillier, S. & Hellings,
L. (2000). Tracing mangrove carbon in suspended matter and aquatic fauna of the
Gautami–Godavari Delta, Bay of Bengal (India). Hydrobiologia 431, 225–241.
DeNiro, M. J. & Epstein, S. (1978). Influence of diet on the distribution of carbon
isotopes in animals. Geochimica et Cosmochimica Acta 45, 341–351.
Dorenbosch, M., Verweij, M. C., Nagelkerken, I., Jiddawi, N. & van der Velde, G.
(2004). Homing and daytime tidal movements of juvenile snappers (Lutjanidae)
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1659
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
between shallow-water nursery habitats in Zanzibar, western Indian Ocean.
Environmental Biology of Fishes 70, 203–209.
Field, A. (2000). Discovering Statistics Using SPSS for Windows. London: Sage Publica-
tions.
Fry, B. & Sherr, E. B. (1984). d13C measurements as indicators of carbon flow on marine
and freshwater ecosystems. Contributions in Marine Science 27, 13–47.
Gearing, J. N. (1991). The study of diet and trophic relationships through natural
abundance 13C. In Carbon Isotope Techniques (Coleman, D. C. & Fry, B., eds), pp.
201–218. San Diego, CA: Academic Press.
Guest, M. A. & Connolly, R. M. (2004). Fine-scale movement and assimilation of
carbon in saltmarsh and mangrove habitats by resident animals. Aquatic Ecology
38, 599–609.
´
Hartill, B. W., Morrison, M. A., Smith, M. D., Boubee, J. & Parsons, D. M. (2003).
Diurnal and tidal movements of snapper (Pagrus auratus, Sparidae) in an estuarine
environment. Marine & Freshwater Research 54, 931–940.
Hobson, K. A. (1999). Tracing origins and migration of wildlife using stable isotopes:
a review. Oecologia 120, 314–326.
Hsieh, H., Chen, C., Chen, Y. & Yang, H. (2002). Diversity of benthic organic matter
flows through polychaetes and crabs in a mangrove estuary: d13C and d34S signals.
Marine Ecology Progress Series 227, 145–155.
Hyslop, E. J. (1980). Stomach content analysis – a review of methods and their
application. Journal of Fish Biology 17, 411–429.
Khalaf, M. A. & Kochzius, M. (2002). Changes in trophic community structure of shore
fishes at an industrial site in the Gulf of Aqaba, Red Sea. Marine Ecology Progress
Series 239, 287–299.
Kieckbusch, D. K., Koch, M. S., Serafy, J. E. & Anderson, W. T. (2004). Trophic
linkages among primary producers and consumers in fringing mangroves of
subtropical lagoons. Bulletin of Marine Science 74, 271–285.
Laegdsgaard, P. & Johnson, C. R. (2001). Why do juvenile fish utilise mangrove habitats?
Journal of Experimental Marine Biology and Ecology 257, 229–253.
Little, M. C., Reay, P. J. & Grove, S. J. (1988). The fish community of an East African
mangrove creek. Journal of Fish Biology 32, 729–747.
Lugendo, B. R., Pronker, A., Cornelissen, I., de Groene, A., Nagelkerken, I.,
Dorenbosch, M., van der Velde, G. & Mgaya, Y. D. (2005). Habitat utilisation
by juveniles of commercially important fish species in a marine embayment in
Zanzibar, Tanzania. Aquatic Living Resources 18, 149–158.
Lugendo, B. R., de Groene, A., Cornelissen, I., Pronker, A., Nagelkerken, I., van der
Velde, G. & Mgaya, Y. D. (in press). Spatial and temporal variation in
fish community structure of a marine embayment in Zanzibar, Tanzania.
Hydrobiologia.
MacDonald, J. S., Waiwood, K. G. & Green, R. H. (1982). Rates of different digestion
of different prey in Atlantic cod (Gadus morhua), ocean pout (Macrozoarces
americanus), winter flounder (Pseudopleuronectes americanus), and American plaice
(Hippoglossoides platessoides). Canadian Journal of Fisheries and Aquatic Sciences
39, 651–659.
Macia, A. (2004). Primary carbon sources for juvenile penaeid shrimps in a mangrove-
fringed bay of Inhaca Island, Mozambique: a dual carbon and nitrogen isotope
analysis. Western Indian Ocean Journal of Marine Science 3, 151–161.
Mahongo, S. B. (1997). Surface heat exchange and bottom reflection in a tidal, shallow
tropical lagoon: Chwaka Bay, Zanzibar. MSc Thesis, Gothenburg University,
Sweden.
Marguillier, S., van der Velde, G., Dehairs, F., Hemminga, M. A. & Rajagopal, S. (1997).
Trophic relationships in an interlinked mangrove-seagrass ecosystem as traced by
d13C an d15N. Marine Ecology Progress Series 151, 115–121.
Minagawa, W. & Wada, E. (1984). Stepwise enrichment of 15N along food chains: further
evidence and the relation between d15N and animal age. Geochimica et Cosmochi-
mica Acta 48, 1135 –1140.
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1660 B. R. LUGENDO ET AL.
Mohammed, S. M., Johnstone, R. W., Widen, B. & Jordelius, E. (2001). The role of
mangroves in the nutrient cycling and productivity of adjacent seagrass commu-
nities, Chwaka Bay, Zanzibar. In Marine Science Development in Tanzania and
Eastern Africa. Proceedings of the 20th Anniversary Conference on Advances in
Marine Science in Tanzania, 1999 (Richmond, M. & Francis, J., eds), pp. 205–226.
Zanzibar: IMS/WIOMSA.
Moncreiff, C. A. & Sullivan, M. J. (2001). Trophic importance of epiphytic algae in
subtropical seagrass beds: evidence from multiple stable isotope analyses. Marine
Ecology Progress Series 215, 93–106.
Nagelkerken, I. & van der Velde, G. (2004). Relative importance of interlinked
mangroves and seagrass beds as feeding habitats for juvenile reef fish on a
Caribbean island. Marine Ecology Progress Series 274, 153–159.
`
Nagelkerken, I., Dorenbosch, M., Verberk, W. C. E. P., Cocheret de la Moriniere, E. &
van der Velde, G. (2000). Day-night shifts 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.
Nakamura, Y. & Sano, M. (2005). Comparison of invertebrate abundance in a seagrass
bed and adjacent coral and sand areas at Amitori Bay, Iriomote Island, Japan.
Fisheries Science 71, 573–550.
Newell, R. I. E., Marshall, N., Sasekumar, A. & Chong, V. C. (1995). Relative
importance of benthic microalgae, phytoplankton, and mangroves as sources of
nutrition for penaeid prawns and other coastal invertebrates from Malaysia.
Marine Biology 123, 595–606.
Nøttestad, L., Giske, J., Chr. Holst, J. & Huse, G. (1999). A length-based hypothesis for
feeding migrations in pelagic fish. Canadian Journal of Fisheries and Aquatic
Sciences 56, 26–34.
Parrish, J. D. (1989). Fish communities of interacting shallow-water habitats in tropical
oceanic regions. Marine Ecology Progress Series 58, 143–160.
Pinnegar, J. K. & Polunin, N. V. C. (1999). Differential fractionation of d13C and d15N
among fish tissues: implication for the study of trophic interactions. Functional
Ecology 13, 225–231.
Polis, G. & Strong, D. R. (1996). Food web complexity and community dynamics.
American Naturalist 147, 813–846.
Quinn, N. J. & Kojis, B. L. (1987). The influence of diel cycle, tidal direction and trawl
alignment on beam trawl catches in an equatorial estuary. Environmental Biology of
Fishes 19, 297–308.
Rao, R. G., Woitchik, A. F., Goeyens, L., van Riet, A., Kazungu, J. & Dehairs, F.
(1994). Carbon, nitrogen contents and stable isotope abundance in mangrove
leaves from an east African coastal lagoon (Kenya). Aquatic Botany 47, 175–183.
Reis, R. R. & Dean, J. M. (1981). Temporal variations in the utilization of an intertidal
creek by the Bay Anchovy (Anchoa mitchilli). Estuaries 4, 16–23.
Robertson, A. I. & Blaber, S. J. M. (1992). Plankton, epibenthos and fish communities.
In Tropical Mangrove Ecosystems (Robertson, A. I. & Alongi, D. M., eds), pp.
173–224. Washington, DC: American Geographical Union.
Robertson, A. I. & Duke, N. C. (1987). Mangrove as nursery sites: comparison of the
abundance and species composition of fish and crustaceans in mangroves and
other near shore habitats in tropical Australia. Marine Biology 96, 193–205.
Rodelli, M. R., Gearing, J. N., Gearing, P. J., Marshall, N. & Sasekumar, A. (1984).
Stable isotope ratios as a tracer of mangrove carbon in Malaysian ecosystems.
Oecologia 61, 326–333.
Rooker, J. R. & Dennis, G. D. (1991). Diel, lunar and seasonal changes in a mangrove
fish assemblage off southwestern Puerto Rico. Bulletin of Marine Science 49,
684–698.
Sheaves, M. & Molony, B. (2000). Short-circuit in the mangrove food chain. Marine
Ecology Progress Series 199, 97–109.
Smith, M. M. & Heemstra, P. C. (Eds) (1991). Smith’s Sea Fishes, 1st edn. Johannesburg:
Southern Book Publishers.
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1661
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
Sydeman, W. J., Hobson, K. A., Pyle, P. & McLaren, E. B. (1997). Trophic relationships
among seabirds in Central California. The Condor 99, 327–336.
de Troch, M., Mees, J. & Wakwabi, E. (1998). Diets of abundant fishes from beach seine
catches in seagrass beds of a tropical bay (Gazi Bay, Kenya). Belgian Journal of
Zoology 128, 135–154.
Vance, D. J., Haywood, M. D. E., Heales, D. S., Kenyon, R. A., Loneragan, N. R. &
Pendrey, R. C. (1996). How far do prawns and fish move into mangroves?
Distribution of juvenile banana prawns, Penaeus merguiensis, and fish in a tropical
mangrove forest in northern Australia. Marine Ecology Progress Series 131,
115–124.
Vanderklift, M. A. & Ponsard, S. (2003). Sources of variation in consumer-diet d15N
enrichment: a meta-analysis. Oecologia 136, 169–182.
van der Veer, H. M. & Bergman, M. N. J. (1986). Development of tidally related
behaviour of a newly settled 0-group plaice (Pleuronectes platessa) populations in
the western Wadden Sea. Marine Ecology Progress Series 31, 121–129.
Verweij, M. C., Nagelkerken, I., de Graaff, D., Peeters, M., Bakker, E. J. & van der
Velde, G. (2006). Structure, food and shade attract juvenile coral reef fish to
mangrove and seagrass habitats: a field experiment. Marine Ecology Progress Series
306, 257–268.
Electronic Reference
Froese, R. & Pauly, D. (Eds) (2004). FishBase. Available at www.fishbase.org (version
08/2004).
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
doi:10.1111/j.1095-8649.2006.01231.x, available online at http://www.blackwell-synergy.com
The importance of mangroves, mud and sand flats, and
seagrass beds as feeding areas for juvenile fishes in
Chwaka Bay, Zanzibar: gut content and stable
isotope analyses
B. R. LUGENDO*†, I. NAGELKERKEN†‡, G. VELDE†§
VAN DER
AND Y. D. MGAYA*
*Faculty of Aquatic Sciences and Technology, University of Dar es Salaam, P. O. Box
35064, Dar es Salaam, Tanzania, †Department of Animal Ecology and Ecophysiology,
Institute for Water and Wetland Research, Faculty of Science, Radboud University,
Toernooiveld 1, 6525 ED Nijmegen, The Netherlands and §National Museum of Natural
History, Naturalis, P.O. Box 9517, 2300 RA Leiden, The Netherlands
(Received 8 December 2005, Accepted 5 July 2006)
The relative importance of bay habitats, consisting of mangrove creeks and channel, seagrass beds,
and mud and sand flats, as feeding grounds for a number of fish species was studied in Chwaka
Bay, Zanzibar, Tanzania, using gut content analysis and stable isotope analysis of carbon and
nitrogen. Gut content analysis revealed that within fish species almost the same food items were
consumed regardless of the different habitats in which they were caught. Crustaceans (mainly
copepods, crabs and shrimps) were the preferred food for most zoobenthivores and omnivores,
while fishes and algae were the preferred food for piscivores and herbivores, respectively. The mean
d13C values of fishes and food items from the mangrove habitats were significantly depleted to
those from the seagrass habitats by 6Á9 and 9Á7% for fishes and food items, respectively, and to
those from the mud and sand flats by 3Á5 and 5Á8%, respectively. Fishes and food items from the
mud and sand flats were significantly depleted as compared to those of the seagrass habitats by 3Á4
and 3Á9%, for fishes and food, respectively. Similar to other studies done in different geographical
locations, the importance of mangrove and seagrass themselves as a primary source of carbon to
higher trophic levels is limited. The different bay habitats were all used as feeding grounds by
different fish species. Individuals of the species Gerres filamentosus, Gerres oyena, Lethrinus lentjan,
Lutjanus fulviflamma, Pelates quadrilineatus and Siganus sutor appeared to show a connectivity with
respect to feeding between different habitats by having d13C values which were in-between those of
food items from two neighbouring habitats. This connectivity could be a result of either daily tidal
migrations or recent ontogenetic migration. # 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles
Key words: feeding areas; habitat connectivity; juvenile fishes; mangroves; stable isotopes;
seagrass beds.
‡Author to whom correspondence should be addressed. Tel.: þ31 24 365 2471; fax: þ31 24 365 2409;
email: i.nagelkerken@science.ru.nl
1639
2006 The Authors
#
Journal compilation # 2006 The Fisheries Society of the British Isles
1640 B. R. LUGENDO ET AL.
INTRODUCTION
Mangrove and seagrass habitats are often characterized by high densities of
juvenile fishes and are therefore often referred to as nursery habitats (Robertson
& Duke, 1987; Little et al., 1988; Parrish, 1989), although little evidence has yet
been provided for this (Beck et al., 2001; Chittaro et al., 2004). Protection
against predation, a high food abundance and easy interception of planktonic
fish larvae due to the large areas of the habitats are among the assumptions
used in explaining the high abundances of juvenile reef fish species in these
habitats (Parrish, 1989; Robertson & Blaber, 1992). Few studies have, how-
ever, tested these hypotheses (Laegdsgaard & Johnson, 2001; Cocheret de la
`
Moriniere et al., 2004; Verweij et al., 2006) in contrast to numerous studies that
describe the fish assemblages of such habitats. The contradicting information
about the functioning of these habitats (Chong et al., 1990) creates a need to
investigate several regions independently. As pointed out by Hartill et al.
(2003), a better understanding is required of the resources used by different fish
species and life stages, and of how important different habitats are in maintain-
ing fish populations before management plans can be improved.
Mangrove and seagrass habitats are often interlinked through diurnal and
tidal fish migrations (Rooker & Dennis, 1991; Vance et al., 1996; Nagelkerken
et al., 2000; Dorenbosch et al., 2004). Little is known, however, of the degree to
which these habitats are used as feeding habitats (Nagelkerken & van der Velde,
2004). Conventional techniques such as gut content analysis may provide unreli-
able results with respect to the diet composition and the source of the food due
to the following reasons: 1) differences in digestion rates of ingested material, 2)
contents can be hard to identify, 3) not all contents are digested, 4) it provides
just a snapshot of the true diet and 5) it does not show from where the food
originates (MacDonald et al., 1982; Gearing, 1991; Polis & Strong, 1996).
Nonetheless, it proves to be the only means of establishing details of the types
and amounts of prey taken (Sydeman et al., 1997). Analysis of the stable iso-
topes of carbon and nitrogen can provide a clearer understanding of diets
because they reflect the actual assimilation of organic matter into consumer tis-
sue rather than merely its consumption, and provide an average of the diet over
periods of weeks to months (Gearing, 1991). The power of stable isotope anal-
ysis as a tool in the investigation of aquatic food web structures and dietary
patterns is based on the significant and consistent differences in isotopic com-
position of different types of primary producers due to different photosynthetic
pathways or different inorganic carbon sources (Bouillon et al., 2002a). The sta-
ble isotopic composition of an animal reflects that of its diet with up to 1Á0%
enrichment in 13C and an average of 3Á5% enrichment in 15N between a consumer
and its food source (DeNiro & Epstein, 1978; Fry & Sherr, 1984; Minagawa &
Wada, 1984) due to the discrimination against lighter isotopes during assimilatory
and excretory functions within consumers (Minagawa & Wada, 1984). The actual
degree of fractionation, however, varies as a function of taxonomy, food quality
and environmental factors (Vanderklift & Ponsard, 2003).
The aim of the present study was to establish the relative importance of dif-
ferent bay habitats, namely, mangroves, seagrass beds, and mud and sand flats,
as feeding areas for juveniles of a number of commercially important fish
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1641
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
species in Chwaka Bay, Zanzibar. The combination of gut analysis and stable
isotope analysis was expected to provide information on both the type and re-
lative amount of prey ingested and to reflect the sources of the food assimilated
by different fish species over periods of weeks up to months. This study endeav-
oured to answer the following questions: 1) Is there a significant difference in
stable isotopic signature (C and N) of fishes and food items in different bay
habitats? 2) In which habitats do fishes eat and what do they consume? 3) To
what degree does connectivity between habitats due to feeding by fishes exist?
MATERIALS AND METHODS
S T U D Y A R EA
The study was carried out in Chwaka Bay, a shallow bay located on the east coast of
Unguja Island, Zanzibar, Tanzania (Fig. 1). Chwaka Bay consists of a large intertidal
flat partly covered with mixed assemblages of algae and seagrass beds with an average
depth of 3Á2 m, an estimated area of 50 km2 at high spring tide and 20 km2 at low
spring tide, and a mean tidal range of 3Á2 m (Cederlof et al., 1995; Mahongo, 1997).
¨
Chwaka Bay is protected from the high-energy ocean on the east coast by a reef system
running along the coastline, as well as the Michamvi Peninsula (Fig. 1). On the land-
ward side, the bay is fringed by a dense mangrove forest of c. 3000 ha (Mohammed
et al., 2001). The mangrove forest has a number of tidal creeks fringed by prop roots
of the mangrove Rhizophora mucronata (Lamarck), with Mapopwe Creek (c. 2 m deep)
being the largest and the main water exchange route between the forest and the bay.
The mangrove creeks and the channel are intertidal in nature and none have any sig-
nificant freshwater input other than rain. The sampled habitats were: mangrove creeks,
mangrove channel, mud and sand flats, Chwaka seagrass beds (seagrass beds close to
the mangroves) and Marumbi seagrass beds (seagrass beds far from mangroves)
(Fig. 1). The sampled seagrass beds consisted of vast fields of Enhalus acoroides (L.)
39°24' 39°30'
N
4 8
0 km
6°6'
Tanzania Zanzibar
5
Marumbi
Mic
ham
Zanzibar
Chwaka
vi P
4
e
nin
Chwaka
3
sula
Bay
6°12' 2
1
FIG. 1. Map of Unguja Island (Zanzibar) showing the location of Chwaka Bay and the sampled habitats
(1, mangrove creeks; 2, mangrove channel; 3, mud and sand flats; 4, Chwaka seagrass beds; 5,
Marumbi seagrass beds). Grey areas in Chwaka Bay indicate mangrove forests.
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1642 B. R. LUGENDO ET AL.
˚
Royle interrupted by small patches of Thalassodendron ciliatum (Forsskal) den Hartog
and the calcareous algae Halimeda spp.
S A M P L IN G D ES I GN
Sample collections were carried out between November 2001 and October 2002. Fish
samples were collected using a seine, while macrofauna and macroflora samples were
collected by hand. Zooplankton samples were collected using a plankton net (80 mm
mesh). In the field, samples were put in a cool box and later frozen at À20° C pending
analysis. Fish species were selected in such a way that they represented commercially
important fish species found abundantly (see Table I) in more than one bay habitat,
and they included five feeding guilds: herbivores [Siganus sutor (Valenciennes)], insecti-
vores [Zenarchopterus dispar (Valenciennes)], omnivores [Monodactylus argenteus (L.)],
piscivores [Sphyraena barracuda (Walbaum)] and zoobenthivores [Gerres filamentosus
˚ `
Cuvier, Gerres oyena (Forsskal), Lethrinus lentjan (Lacepede), Lutjanus fulviflamma
˚
(Forsskal) and Pelates quadrilineatus (Bloch)]. Fish guild membership was assigned using
Smith & Heemstra (1991), Khalaf & Kochzius (2002) and Froese & Pauly (2004), which
were also used as a guide for the sampling of potential food items for each fish species.
Detailed information on the environmental variables and the fish community structure
(and their temporal variation) of Chwaka Bay can be found in other studies (Lugendo
et al., 2005, in press; B. R. Lugendo, I. Nagelkerken, N. S. Jiddawi, G. van der Velde
and Y. D. Mgaya, unpubl. data).
S T A B L E I S O T O PE A N A L Y SI S
Muscle tissues were removed from the fishes, while molluscs (gastropods and bi-
valves) and crustaceans (crabs and shrimps) were dissected from their exoskeleton
or shells prior to drying. The zooplankton samples were cleaned from detritus, sedi-
ments and other materials, under a dissecting microscope. Samples were dried at 70° C
for 48 h and ground to powder (homogeneous mixture). For samples rich in carbo-
nates such as detritus and whole individuals of small hermit crabs, sub-samples were
acid-washed and oven-dried. These sub-samples were used for stable carbon isotope
analysis only, while the remaining untreated sub-samples were used for stable nitrogen
isotope analysis since acid-washing interferes with stable nitrogen isotopes (Pinnegar &
Polunin, 1999). Samples were placed in ultra-pure tin capsules and combusted in a Carlo
ErbaÒ NA 1500 elemental analyser coupled on-line via a Finnigan Conflo III interface
with a ThermoFinnigan DeltaPlus mass spectrometer. Carbon and nitrogen isotope
ratios are expressed in the delta notation (d13C and d15N) relative to Vienna PDB
and atmospheric nitrogen. The potential food items and possible feeding habitat for
fishes were determined in view of the enrichment in isotope signatures of 1 and 3Á5%,
for carbon and nitrogen, respectively, between fishes and their potential food items
(DeNiro & Epstein, 1978; Minagawa & Wada, 1984). The term ‘macroinvertebrate’ is
used in the figures to denote zoobenthos and insects together, while the term ‘zoobenthos’
whenever used in the figures excludes the insects.
GUT CONTENT ANALYSIS
For fishes, fork length (LF) was measured to the nearest 0Á1 cm, and the entire gut
extracted and frozen pending analysis. The gut was then split, the gut contents placed
in a Petri dish under a dissecting microscope and food items were identified to the low-
est taxa possible. The percentage of the total stomach volume that each food category
comprised was determined using the point method (Hyslop, 1980) in which the food
items in each fish gut was allotted a number of points depending on its abundance
and size of an organism (i.e. one large organism counted as much as a large number
of small ones). The points and the percentages they represented were 5 (75–100%), 4
(50–75%), 3 (25–50%), 2 (5–25%) and 1 (up to 5%). All the points gained by each
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
TABLE I. Stable carbon and nitrogen isotope signatures (mean Æ S.E.) of different fork length (LF) classes of fish species in different bay
habitats. The overall mean d13C is also shown where more than one size class of fish species was present in a habitat. Relative abundance and
relative biomass for each species in each bay habitat are also given. Numbers in bold print show relative proportions of each species for the
whole bay. Different superscript lowercase letters and numbers represent statistical post hoc results and denote significantly different
(P < 0Á05) stable carbon isotope values of a fish species for similar LF classes and for overall d13C among different bay habitats
# 2006 The Authors
Overall Relative Relative
LF class
N
Species (cm) d13C d15N Species mean d13C abundance (%) biomass (%)
Gerres filamentosus Gerres filamentosus 7Á5 3Á6
Mangrove creeks 0–5 2 0Á3 8Á3 0Á1 Mangrove creeks 9Á6 5Á8
À21Á3 Æ Æ À21Á2 Æ 0Á4a
Mangrove creeks 5–10 3 0Á5a 8Á0 0Á1 Mangrove channel 24Á3 17Á0
À21Á1 Æ Æ À21Á6 Æ 0Á5a
Mangrove channel 5–10 5 0Á5a 8Á4 0Á1 Mud and sand flats 0Á8 1Á1
À21Á6 Æ Æ À19Á2 Æ 1Á0a
Mud and sand flats 5–10 4 1Á0a 8Á0 0Á2
À19Á2 Æ Æ
Gerres oyena Gerres oyena 22Á6 21Á2
Mangrove creeks 5–10 10 0Á3a 7Á5 0Á2 Mangrove creeks 7Á1 7Á1
À19Á4 Æ Æ
Mangrove channel 5–10 10 0Á7b 7Á4 0Á1 Mangrove channel 25Á6 35Á0
À17Á0 Æ Æ
Mud and sand flats 5–10 10 0Á3c 6Á6 0Á2 Mud and sand flats 62Á6 56Á2
À13Á8 Æ Æ
Chwaka seagrass beds 5–10 10 0Á7c 7Á5 0Á1 Chwaka seagrass beds 37Á6 38Á8
À12Á8 Æ Æ
Lethrinus lentjan Lethrinus lentjan 2Á7 1Á6
Mangrove channel 5–10 10 0Á3a 8Á0 0Á1 Mangrove channel 0Á3a 2Á4 1Á2
À21Á8 Æ Æ À21Á8 Æ
Mud and sand flats 5–10 9 0Á7b 6Á8 0Á2 Mud and sand flats 0Á7b 6Á8 3Á4
À19Á3 Æ Æ À19Á3 Æ
Chwaka seagrass beds 5–10 10 0Á2c 8Á0 0Á1 Chwaka seagrass beds 0Á2c 3Á9 3Á0
À12Á3 Æ Æ À12Á3 Æ
Marumbi seagrass beds 5–10 4 0Á6c 8Á3 0Á2 Marumbi seagrass beds 0Á4c 1Á7 1Á1
À12Á4 Æ Æ À12Á0 Æ
Marumbi seagrass beds 10–15 2 0Á1 8Á3 0Á0
À11Á6 Æ Æ
Lutjanus fulviflamma Lutjanus fulviflamma 2Á0 3Á3
Mangrove channel 5–10 4 0Á21 8Á5 0Á1 Mangrove creeks 0Á8 2Á0
À21Á0 Æ Æ À20Á1 Æ 0Á9a
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
Chwaka seagrass beds 5–10 4 0Á72 8Á0 0Á2 Mangrove channel 1Á4 2Á1
À15Á2 Æ Æ À21Á8 Æ 0Á1a
Mangrove creeks 10–15 3 0Á9ab 8Á6 0Á4 Mud and sand flats 3Á8 5Á6
À20Á1 Æ Æ À15Á2 Æ 0Á5b
Mangrove channel 10–15 2 0Á0a 9Á2 0Á0
À22Á6 Æ Æ
Mud and sand flats 10–15 5 0Á5c 7Á6 0Á2
À15Á2 Æ Æ
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1643
TABLE I. Continued
1644
Overall Relative Relative
LF class
N
Species (cm) d13C d15N Species mean d13C abundance (%) biomass (%)
Chwaka seagrass beds 10–15 5Á2 8Á2
2 À14Á2 Æ 0Á0bc 9Á0 Æ 0Á2 Chwaka seagrass beds À14Á5 Æ 0Á4b
Marumbi seagrass beds 10–15 0Á3 0Á4
Journal compilation
4 À11Á2 Æ 0Á4d 9Á0 Æ 0Á1 Marumbi seagrass beds À11Á2 Æ 0Á4c
#
Monodactylus argenteus Monodactylus argenteus 3Á1 1Á2
Mangrove creeks 0–5 5Á5 2Á8
11 À22Á1 Æ 0Á2a 8Á0 Æ 0Á1 Mangrove creeks
Mangrove channel 5–10 5Á5 3Á8
3 À24Á0 Æ 0Á6b 8Á4 Æ 0Á4 Mangrove channel
Pelates quadrilineatus Pelates quadrilineatus 2Á5 2Á4
Mud and sand flats 5–10 2Á9 1Á9
9 À17Á1 Æ 0Á1a 7Á2 Æ 0Á1 Mud and sand flats
Chwaka seagrass beds 5–10 12Á3 10Á8
10 À16Á2 Æ 0Á4b 7Á8 Æ 0Á2 Chwaka seagrass beds
Siganus sutor Siganus sutor 1Á6 3Á6
Mud and sand flats 5–10 7 0Á5a 1Á3 1Á2
À22Á8 Æ 5Á6 Æ 0Á3 Mud and sand flats À22Á8 Æ 0Á5a
Chwaka seagrass beds 5–10 4 0Á8a 2Á8 2Á0
À20Á7 Æ 7Á0 Æ 0Á2 Chwaka seagrass beds À19Á5 Æ 0Á7b
Marumbi seagrass beds 5–10 4 0Á7b 7Á4 11Á4
À15Á5 Æ 6Á7 Æ 0Á3 Marumbi seagrass beds À16Á1 Æ 0Á5c
Chwaka seagrass beds 10–15 1 6Á1
À15Á4
Marumbi seagrass beds 10–15 11 À16Á2 Æ 0Á6 6Á5 Æ 0Á1
Marumbi seagrass beds 15–20 11 À16Á5 Æ 0Á2 6Á3 Æ 0Á1
B. R. LUGENDO ET AL.
Sphyraena barracuda Sphyraena barracuda 0Á9 3Á8
Mangrove creeks 10–15 2 1Á01 1Á0a 0Á9 3Á1
À20Á6 Æ 9Á2 Æ 0Á0 Mangrove creeks À20Á6 Æ
Mangrove channel 10–15 5 0Á51 0Á5a 1Á1 4Á5
À19Á9 Æ 9Á8 Æ 0Á1 Mangrove channel À19Á9 Æ
Mud and sand flats 15–20 5 0Á2 0Á4b 1Á5 7Á2
À15Á7 Æ 8Á5 Æ 0Á2 Mud and flats À15Á9 Æ
Mud and sand flats 20–25 2 0Á5a 1Á8b 0Á7 5Á9
À16Á1 Æ 8Á2 Æ 0Á4 Chwaka seagrass beds À14Á6 Æ
Chwaka seagrass beds 20–25 2 1Á8a
À14Á6 Æ 9Á1 Æ 0Á6
Zenarchopterus dispar Zenarchopterus dispar 3Á9 4Á8
Mangrove creeks 10–15 8Á2 14Á6
9 À22Á8 Æ 0Á1a 8Á1 Æ 0Á1 Mangrove creeks
Mangrove channel 10–15 2Á9 4Á9
9 À22Á7 Æ 0Á1a 8Á2 Æ 0Á1 Mangrove channel
N, sample size.
2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
# 2006 The Authors
1645
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
food item were scaled down to percentages, to give percentage composition of each
food item in a diet of individual fish species examined.
STATISTICAL ANALYSIS
Each bay habitat was treated as a sample unit. First, data were pooled for each hab-
itat for fishes and for food items, respectively, in order to test for the overall differences
among habitats. Subsequently, each fish species was treated separately. The numbers of
individual fishes analysed for each particular species (i.e. sample size) equalled the num-
ber of replicates (N; Table I). Data were checked for homogeneity of variances using
a Levene’s test (Field, 2000). In case variances were homogeneous, a one-way ANOVA
or t-test was employed to test for differences in stable isotope signatures of carbon for
fishes and food items among different habitats. Since fish sample sizes were very differ-
ent (see Table I), a Hochberg’s GT2 was used as a post hoc test due to its greater sta-
tistical power in such kinds of data compared to other tests (Field, 2000). All data that
did not show homogeneous variances were log10-transformed, and a Levene’s test was
performed once again. Either Kruskal–Wallis test or Mann–Whitney U-test (depending
on the number of sample units involved) on the non-transformed data was used as a
non-parametric test equivalent when variances were not homogeneous, even after log10-
transformation. A Games–Howell post hoc test was used following the Kruskal–Wallis
tests because it is more powerful and specifically designed for lack of homogeneity of var-
iances (Field, 2000). A significance level of P < 0Á05 was used in all tests. All analyses
were performed using the programme SPSS 11.5 for Windows (Field, 2000).
RESULTS
GUT CONTENT ANALYSIS
Gut analysis indicated a food preference by different fish species, despite the
fact that they ingested a variety of food items (Table II). While some fish spe-
cies maintained a quite similar diet type regardless of the different habitats
from which they were caught (G. filamentosus: copepods; S. sutor: macroalgae;
S. barracuda: fishes; Z. dispar: insects), the diet of the other species (G. oyena,
L. lentjan, L. fulviflamma and M. argenteus) differed within species in different hab-
itats. The main food of G. oyena from the mangrove channel and from Chwaka
seagrass beds consisted mainly of copepods while fishes from mud and sand
flats fed mainly on detritus (Table II). Lethrinus lentjan fed mainly on ostracods
in the mangrove channel, on copepods on the mud and sand flats and on crus-
taceans and insects in the Chwaka seagrass beds. The diet of L. fulviflamma
consisted mainly of crustaceans in the mangroves, of copepods on the mud
and sand flats, of crabs and shrimps in Chwaka seagrass beds, and of crabs
and fishes in Marumbi seagrass beds. Monodactylus argenteus from the man-
grove creeks fed mainly on copepods while those from mangrove channel fed
mainly on algae (Table II).
M E A N d1 3 C S I G N A T U R E S F O R F I S H E S A N D F O O D I T E M S
A clear gradient in d13C could be discerned for fishes as well as food items
from the mangrove habitats located deep into the bay to the seagrass beds at
the mouth of the bay (Fig. 2). Fishes and food items from the mangrove hab-
itats were significantly depleted (Hochberg’s GT2, P < 0Á001) to those from the
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
TABLE II. Mean percentage composition of diet for different fish species and fork length (LF) classes in different bay habitats. Grey boxes
highlight all food items with a relative abundance of >19%.
1646
Unidentified Unidentified
Crustacean animal plant
LF class
Species/site (cm) N Copepod Crab Shrimp Ostracod Parts Fishes Detritus Gastropod Nematode Insect Algae Seagrass Sediment material material Other
Journal compilation
Gerres filamentosus (ZB)
#
Mangrove 0–5, 5–10 3 95Á8 1Á0 3Á2
creeks
Mangrove 0–5, 5–10 45 71Á2 1Á9 0Á3 11Á4 0Á1 2Á3 9Á9 2Á8
channel
Gerres oyena (ZB)
Mangrove 5–10 15 42Á9 2Á0 15Á2 7Á1 0Á2 9Á8 12Á5 6Á2 4Á2
channel
Mud and 0–5, 5–10, 16 21Á2 2Á1 53Á3 8Á4 0Á2 1Á9 8Á7 4Á2
sand flats 10–15
Chwaka 10–15 11 39Á2 2Á7 26Á1 3Á4 4Á8 11Á2 12Á6
seagrass beds
Lethrinus lentjan (ZB)
Mangrove 5–10 7 12Á9 53Á6 14Á3 2Á1 17Á1
channel
Mud and 5–10 12 70Á8 3Á1 8Á3 0Á3 0Á3 17Á0 0Á3
sand flats
Chwaka 5–10, 10–15 3 33Á3 33Á3 33Á3
B. R. LUGENDO ET AL.
seagrass beds
Lutjanus fulviflamma (ZB)
Mangrove 5–10 6 19Á2 19Á2 27Á1 16Á7 16Á7 1Á2
channel
Mud and 5–10, 10–15 12 39Á6 8Á3 20Á8 0Á3 8Á3 8Á3 14Á3
sand flats
Chwaka 5–10, 10–15 30 40Á4 22Á7 10Á4 0Á1 4Á2 20Á1 2Á1
seagrass beds
Marumbi 10–15 3 45Á8 20Á8 33Á3
seagrass beds
Monodactylus argenteus (O)
Mangrove 0–5, 5–10 18 46Á0 50* 0Á8 3Á0 0Á2
creeks
2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
# 2006 The Authors
TABLE II. Continued
Unidentified Unidentified
# 2006 The Authors
Crustacean animal plant
LF class
Species/site (cm) N Copepod Crab Shrimp Ostracod Parts Fishes Detritus Gastropod Nematode Insect Algae Seagrass Sediment material material Other
Mangrove 5–10 10 6Á3 17Á8 10Á0 65Á0 0Á9
channel
Pelates quadrilineatus (ZB)
Mud and 5–10, 10–15 5 32Á5 27Á5 20Á0 20Á0
sand flats
Zenarchopterus dispar (ZB)
Mangrove 10–15, 15–20 9 0Á3 11Á4 9Á7 69Á4 4Á2 4Á2 0Á7
creeks
Mangrove 15–20 5 20Á0 20Á0 40Á0 20Á0
channel
Siganus sutor (H)
Mangrove 5–10 5 92Á5 7Á5
channel
Mud and 5–10, 10–15 5 4Á4 72Á5 7Á5 15Á6
sand flats
Marumbi 5–10, 10–15, 23 7Á1 79Á4 2Á3 11Á2
seagrass beds 15–20
Sphyraena barracuda (P)
Mangrove 10–15 5 100Á0
channel
Mud and 10–15, 15–20, 7 16Á7 62Á5 1Á0 18Á8 1Á0
sand flats 20–25
Chwaka 10–20, 20–30 4 21Á9 1Á0 71Á9 1Á0 1Á5 2Á7
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
seagrass beds
H, herbivore; O, omnivore; P, piscivore; ZB, zoobenthivore; N, number of fish analysed; *, shrimp larvae.
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1647
1648 B. R. LUGENDO ET AL.
seagrass habitats by on average 6Á9 and 9Á7% for fishes and food items, respec-
tively, and from the mud and sand flats by on average 3Á5 and 5Á8%, respec-
tively. Food items from the mud and sand flats were significantly depleted
(Hochberg’s GT2, P < 0Á01) as compared to those of the seagrass habitats
by on average 3Á9%. Fishes from the mud and sand flats were depleted by
an average of 3Á4%, but this difference was not significant (Hochberg’s GT2,
P > 0Á05). There were no significant differences (Hochberg’s GT2, P > 0Á05)
in d13C between the two mangrove habitats (average difference of 0Á2 and
1Á5%, for fishes and food, respectively), and between the two seagrass habitats
(average difference of 1Á9 and 0Á2%, for fishes and food, respectively).
TROPHIC LEVELS OF FISHES AND FOOD ITEMS
The food web in the bay showed various trophic levels. Detritus and plant
material were generally more depleted in d15N as compared to zooplankton
and macroinvertebrates (zoobenthos þ insects) found within the same habitat,
while fishes were the most enriched in d15N (Fig. 3). Also, clear gradients in
both d13C and d15N could be observed for different feeding guilds of fishes
and for different habitats [Fig. 3(b)]. Three trophic levels could be discerned
for the fishes, with increasing values of d15N from herbivores to omnivores
and zoobenthivores to piscivores. For each feeding guild, d13C increased along
the spatial gradient from mangroves in the bay to seagrass beds at the mouth
of the bay [Fig. 3(b)].
S T A B L E I S O T O PI C S I G N A T U R E S O F F I SH S PE C I E S
Individual fish species from the mangrove habitats were generally more
depleted in d13C compared to those of the same species from either mud and
9
8
7
6
N
15
5
4
3
2
–26 –25 –24 –23 –22 –21 –20 –19 –18 –17 –16 –15 –14 –13 –12
13
C
FIG. 2. Pooled mean Æ S.E. stable carbon and nitrogen isotope values of fishes (), u, n, s, *) and food
items (r, n, m, d, Â) in different bay habitats. (), r, mangrove creeks; u, n, mangrove channel;
n, m, mud and sand flats; s, d, Chwaka seagrass beds; *, Â, Marumbi seagrass beds).
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1649
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
(a)
7
6
Zoobenthos
5
Zoobenthos
4 Macro-
Mangroves + Zooplankton Zooplankton
invertebrates
seagrasses
Zoobenthos
Zooplankton
Detritus
Detritus
Algae +
3 Zooplankton Macro- seagrasses
Zooplankton
Algae +
invertebrates
seagrasses
Algae +
2 seagrasses
Detritus
Detritus
Algae +
1 seagrasses
0
–28 –26 –24 –22 –20 –18 –16 –14 –12
N
15
13
C
(b)
11
Mangrove channel
10
Mangrove Chwaka
creeks seagrass beds Marumbi
Mangrove
9 seagrass beds
channel
Mangrove creeks
Mud and sand flats
8 Mangrove
Mangrove
Mud and sand flats
channel Chwaka
creeks
seagrass beds
7
Chwaka
seagrass beds Marumbi
6 seagrass beds
Mud/sand flats
5
–25 –23 –21 –19 –17 –15 –13 –11
13
C
FIG. 3. Mean d13C and d15N values of different (a) trophic groups of food items in different bay habitats.
(), mangrove creeks; u, mangrove channel; n, mud and sand flats; s, Chwaka seagrass beds; *,
Marumbi seagrass beds), and (b) feeding guilds of fish in different bay habitats (n, piscivores; u,
omnivores; ), zoobenthivores; s, herbivores).
sand flats or the seagrass habitats (Table I). The d13C depletion of individual
species was generally in the order: mangrove habitats < mud and sand flats
< seagrass habitats. The highest enrichment in d13C between two neighbouring
habitats was observed for individuals of the same LF class (10–15 cm) of L.
fulviflamma (mangrove channel and mud and sand flats: 7Á4%). With regard
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1650 B. R. LUGENDO ET AL.
to d15N, the herbivore S. sutor was the most depleted and the piscivore S. bar-
racuda the most enriched fish species, with a range of 2Á6–2Á8% (considering
the overall mean for LF classes) between the two species when occurring in the
same habitat. Considering individuals of the same species and similar size classes
in different bay habitats, d13C of five species, namely, G. oyena (5–10 cm), L.
lentjan (5–10 cm), L. fulviflamma (10–15 cm), P. quadrilineatus (5–10 cm) and
S. sutor (5–10 cm), differed significantly between habitats (one-way ANOVA,
d.f. ¼ 3,36, P < 0Á001 for G. oyena, Kruskal–Wallis, d.f. ¼ 3, P < 0Á001 for
L. lentjan, Kruskal–Wallis, d.f. ¼ 4, P < 0Á01 for L. fulviflamma, t-test, d.f. ¼ 1,
P < 0Á05 for P. quadrilineatus and one-way ANOVA, d.f. ¼ 2,12, P < 0Á001
for S. sutor), while those of G. filamentosus (5–10 cm) and Z. dispar (10–15 cm)
did not differ significantly between different bay habitats (Kruskal–Wallis,
d.f. ¼ 2, P > 0Á05 for G. filamentosus, Mann–Whitney U-test, d.f. ¼ 1, P >
0Á05 for Z. dispar). Similar results were obtained when different LF classes of indi-
vidual species were pooled within each habitat in which case also M. argenteus
differed significantly between the two mangrove habitats (t-test, d.f. ¼ 1, P <
0Á01). The post hoc results are presented in Table I.
ANALYSIS OF POTENTIAL FOOD AND FEEDING
H A B I T A T S O F D I F F E R E N T F I S H S P E C I ES
The herbivore S. sutor ingested mainly macroalgae (Table II). The d13C and
15
d N values of the average diet of this species were generally quite similar to
those of macroalgae, but very distinct from those of seagrasses or mangrove
leaves [Fig. 4(a)]. Siganus sutor from the mud and sand flats showed stable iso-
tope signatures indicating various types of macroalgae from the mangrove chan-
nel as a potential food source, while fish from Chwaka seagrass beds showed
values indicating green and brown algae from mud and sand flats and the cal-
careous green algae (Halimeda sp.) from Chwaka seagrass beds as a potential
food source. Siganus sutor from Marumbi seagrass beds showed an intermediate
value for its average diet that lay in-between those of green algae from the mud
and sand flats, Halimeda sp. from Chwaka seagrass beds, and calcareous green
algae (Udotea sp.) and red algae from Marumbi seagrass beds.
The gut content of the insectivore Z. dispar showed that insects formed a major
part of its diet (Table II), while the stable isotope values from both mangrove
habitats suggested a mixed diet of crabs (Sesarma sp. and Portunidae), shrimps
and insects from the mangroves [Fig. 4(b)].
FIG. 4. Mean Æ S.E. d13C and d15N values of fish species (a) Siganus sutor, (b) Zenarchopterus dispar, (c)
Monodactylus argenteus, (d) Sphyraena barracuda, (e) Gerres filamentosus, (f) Gerres oyena, (g)
Lethrinus lentjan, (h) Lutjanus fulviflamma and (i) Pelates quadrilineatus (large symbols) and potential
food items (small symbols) in different bay habitats. (), mangrove creeks; u, mangrove channel; n,
mud and sand flats; s, Chwaka seagrass beds; *, Marumbi seagrass beds). The arrow heads indicate
the predicted average d13C and d15N values (based on the 1 and 3Á5% enrichment, respectively, in
d13C and d15N between an animal and its food source) of the diet of fishes. The dashed lines combine
potential food sources within a habitat. Prey species are depicted on lowest taxonomic level for each
habitat in which the fish species was found; for the remainder of the habitats the prey species are
pooled to higher taxonomic levels (e.g. macroinvertebrates, zoobenthos and seagrasses).
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1651
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
(a)
8
6
Red algae
Red algae
Green algae
Mangrove
Thalassia
Green
Green
leaves
4 hemprichii
algae
algae Enhalus
Brown
Brown Udotea sp. acoroides
algae
Green algae
Red algae Thalassodendron
Red algae algae
ciliatum
Thalassodendron
Halimeda sp. Thalassia
ciliatum
2 Eucheuma sp. hemprichii
Halimeda Thalassia
Jania sp.
sp. hemprichii
Halimeda
Halodule sp. Enhalus acoroides Enhalus
wrightii
Cymodocea sp. acoroides
0
Brown
algae
Halodule
wrightii
–2
–30 –28 –26 –24 –22 –20 –18 –16 –14 –12 –10 –8 –6
(b)
9
7 Insects
Insects
Sesarma sp.
5 Animals
Shrimps
N
Portunidae Animals
15
Terebralia sp.
Red algae
Zooplankton
Detritus Brown algae Uca spp. Animals
3 Zooplankton
Algae +
seagrasses Algae +
Green algae
Detritus seagrasses
Cyanobacteria
1
Algae +
Seagrasses
Halodule wrightii seagrasses
–1
–30 –28 –26 –24 –22 –20 –18 –16 –14 –12 –10
(c)
10
8
Insects
Sesarma Insects Zoobenthos
6 sp.
Shrimps Zooplankton
Zoobenthos
Portunidae
4 Terebralia sp. Zooplankton
Red algae
Zooplankton Brown Zoobenthos
Detritus
algae
Detritus Algae +
Green algae Algae +
seagrasses
Zooplankton seagrasses
Zooplankton
Uca
2 spp.
Algae +
Halimeda sp.
seagrasses
Detritus Detritus
Thalassia
hemprichii
Halodule wrightii
0
Halodule wrightii
–2
–30 –28 –26 –24 –22 –20 –18 –16 –14 –12 –10
13
C
FIG. 4. Continued
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1652 B. R. LUGENDO ET AL.
11 (d)
10
9
Omnivores Carnivores
Carnivores
8 Carnivores
Omnivores Carnivores
Carnivores
7 Herbivores
?
Herbivores
6 ?
Herbivores
? ? Zoobenthos
?
?
?
5
Zoobenthos
4 Zoobenthos
Zoobenthos
Zoobenthos
3
–25 –23 –21 –19 –17 –15 –13 –11
9 (e)
8
7
Benthic
invertebrates
Bivalve sp.1
6 Sesarma sp.
N
Gastropods
Shrimps
5 Polychaetes
15
Benthic invertebrates
Zooplankton Terebralia sp.
4 Portunidae
Shrimps
Zooplankton
Terebralia sp.
Zooplankton
3 Zooplankton
Zooplankton Bivalve sp.2
Uca spp.
Hermit crabs
2 Dotilla
Crab sp. fenestrata
1
–28 –26 –24 –22 –20 –18 –16 –14 –12 –10
9 (f)
8
7 Isopods
Cypraea sp.
Brittle stars
6 Polychaetes
Sesarma sp. Bivalve sp.1
Thalamita sp.
Gastropods
Shrimps Zooplankton
5 Polychaetes
Zoobenthos
Amphipods
Portunidae
Shrimps
4
Hermit crabs
Zooplankton
Zooplankton Terebralia sp.
Terebralia sp. Dotilla
3 fenestrata
Zooplankton Zooplankton
Uca spp. Bivalve sp.2
Hermit crabs
2
Crabs
Dotilla
1 fenestrata
0
–26 –24 –22 –20 –18 –16 –14 –12 –10
13
C
FIG. 4. Continued
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1653
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
10 (g)
8 Littoraria sp.
Brittle stars
Isopods
Cypraea sp.
Polychaetes
Brittle stars
6 Sesarma sp. Bivalve sp.1
Thalamita sp.
Gastropods Polychaetes
Shrimps
Zooplankton Pinna sp.
Amphipods
Polychaetes
Shrimps
4 Hermit crabs
Portunidae Terebralia sp. Zooplankton Dotilla fenestrata
Terebralia sp. Hermit crabs
Zooplankton Zooplankton
Zooplankton Uca spp. Gastropods
Bivalve sp.2
Hermit
2 crabs
Crabs
Dotilla
fenestrata
0
–28 –26 –24 –22 –20 –18 –16 –14 –12 –10 –8
10 (h)
8 Littoraria sp.
Isopods Cypraea sp.
Brittle stars
Polychaetes
Brittle stars
6 Sesarma sp.
Thalamita sp.
Polychaetes
N
Pinna sp. Gastropods
15
Zooplankton
Polychaetes Amphipods
Shrimps
Portunidae Shrimps
Hermit crabs
4 Bivalves
Zooplankton Dotilla
fenestrata
Zooplankton Hermit
Terebralia sp. Terebralia sp.
crabs
Zooplankton
Zooplankton Zooplankton
Dotilla
Uca spp.
2 Hermit crabs fenestrata
Crabs
Dotilla
fenestrata
0
–28 –26 –24 –22 –20 –18 –16 –14 –12 –10 –8
(i)
9
8
7 Brittle stars
Brittle stars
6
Zoobenthos
5 Pinna sp.
Zooplankton
Gastropod sp.
4 Zooplankton Zoobenthos
Zooplankton Zoobenthos
3 Zoobenthos
Zooplankton
Zooplankton
Zoobenthos
Hermit crabs
2
1
0
–28 –26 –24 –22 –20 –18 –16 –14 –12 –10
13
C
FIG. 4. Continued
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1654 B. R. LUGENDO ET AL.
The omnivore M. argenteus ingested zooplankton, algae and some detritus
(Table II). The ingestion of zooplankton and detritus is supported by the
d13C for fish from the mangrove channel, although the enrichment in d15N
was larger than the usual 3Á5% [Fig. 4(c)]. For fish from the mangrove creeks,
the d13C signature suggests the diet to consist of a mixture of decapods (Sesarma
sp., Portunidae and shrimps) from the mangrove creeks and zooplankton and
detritus from the mangrove channel, but without an indication of dependence
on algae as a food source [Fig. 4(c)].
Although the gut content analysis shows that fishes formed major part of the
diet of the piscivore S. barracuda, the stable isotope signatures of the average
diet of S. barracuda was not close enough to those of the selected fish species
of this study to depend solely on these species as a food source. Sphyraena
barracuda from the mangrove habitats had an isotope signature of its average
diet that was closest to that of herbivorous fishes from the mud and sand flats
and Chwaka seagrass beds, while for S. barracuda from the mud and sand flats
and Chwaka seagrass beds this was the case for herbivorous fish from the Mar-
umbi seagrass beds, with a possibility of feeding partly on macrofauna too
[Fig. 4(d)].
In conformity with the gut content analysis where crustaceans (mainly cope-
pods, crabs and shrimps) formed a major part of the diet of most zoobenthi-
vores (Table II), G. filamentosus from the mangrove habitats had stable
isotope signatures for its average diet which lay in-between those of crustaceans
(Sesarma sp., Portunidae and shrimps) from the mangrove creeks, while fish
from mud and sand flats had isotope signatures for their average diet which
lay in-between values for shrimps from the mangrove creeks and zooplankton
from mud and sand flats [Fig. 4(e)]. Gerres oyena from the mangrove creeks
showed an isotope signature of its average diet close to the signatures of
shrimps from the mangrove creeks, gastropods (Terebralia sp.) from the man-
grove channel and zooplankton from the mud and sand flats, while G. oyena
from mangrove channel had signatures closest to zooplankton from the mud
and sand flats [Fig. 4(f)]. Gerres oyena from the mud and sand flats showed
an average diet signature close to that of bivalves, gastropod (Terebralia sp.)
and shrimps from mud and sand flats and zooplankton from the seagrass beds.
Gerres oyena from Chwaka seagrass beds showed a signature of its average diet
close to that of shrimps and gastropods (Terebralia sp.) from the mud and sand
flats, hermit crabs and amphipods from the Chwaka seagrass beds, and zoo-
benthos from the Marumbi seagrass beds [Fig. 4(f)]. Lethrinus lentjan from
the mangrove channel showed an isotope signature of its average diet that
was intermediate between crabs and shrimps of the mangrove creeks, while
for the mud and sand flats the signatures suggested a possible mix of shrimps,
crabs (Uca spp.) and gastropod (Terebralia sp.) from the mangroves and hermit
crabs and zooplankton from the mud and sand flats as a food source [Fig. 4(g)].
Lethrinus lentjan from the seagrass habitats showed an average stable iso-
tope signature for its diet that was close to that of the zoobenthos from the
seagrass habitats. The isotope signature of the average diet of L. fulviflamma
from the mangrove habitats showed proximity to isotope signatures of crabs
and shrimps from the mangrove habitats, while that of fish from the mud
and sand flats and Chwaka seagrass beds suggested an intermediate isotope
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1655
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
signature of polychaetes, shrimps and zooplankton from the mud and sand flats
[Fig. 4(h)]. Lutjanus fulviflamma from the Marumbi seagrass bed showed a stable
isotope signature of its diet in close proximity to zoobenthos from the seagrass
beds. Pelates quadrilineatus from both the mud and sand flats and Chwaka sea-
grass beds showed isotope signatures for their average diet close to zooplankton
from mud and sand flats and the two seagrass beds [Fig. 4(i)].
C O N N E C T I V I T Y B ET W E E N H A B I T A T S
The isotopic signatures of the fish species in relation to that of the possible
food items suggest four possibilities of feeding connectivity between adjacent
bay habitats (Fig. 4): 1) connectivity between the two mangrove habitats for
G. filamentosus, L. lentjan, L. fulviflamma, M. argenteus and Z. dispar, 2) connec-
tivity between mangrove habitats and mud and sand flats for G. filamentosus,
L. lentjan and S. sutor, 3) connectivity between mud and sand flats and seagrass
habitats for G. oyena, L. fulviflamma, P. quadrilineatus and S. sutor, and 4) con-
nectivity between the two seagrass habitats for L. fulviflamma and L. lentjan.
DISCUSSION
Both gut content and stable carbon isotope analyses showed evidence that
the studied fish species generally relied as a food source on algae (herbivores)
and macroinvertebrates (omnivores and zoobenthivores), with crustaceans
(crabs, shrimps and copepods) playing a major role. The different d13C or
d15N values of the piscivore S. barracuda from those of herbivorous fishes
indicate a possible dependence for juveniles (10–25 cm) of this species on other
animals than fishes alone. Copepods were found to some degree in the guts of
the juveniles. In a study in Gazi Bay (Kenya), de Troch et al. (1998) identified
other animals like gammaridean amphipods, mysids, crabs and shrimps in the
stomachs of piscivorous fishes (including S. barracuda), an observation that in-
dicates that at juvenile stages S. barracuda is not solely piscivorous.
Although the stable isotope signatures showed evidence for food dependence
of the studied fish species on mangrove and seagrass habitats, the direct con-
sumption of either mangrove or seagrass leaves seemed to be absent or very
low. The mean d13C of mangrove leaves of À28Á1% is similar to the overall
values for mangrove leaves recorded in the Caribbean, India, Malaysia and
in Kenya (Rao et al., 1994; Chong et al., 2001; Bouillon et al., 2002a; Cocheret
`
de la Moriniere et al., 2003). Similar to what was observed by Sheaves & Molony
(2000), Bouillon et al. (2002b) and Kieckbusch et al. (2004), however, this
value is much more depleted as compared to either fish species (Sheaves &
Molony, 2000; Kieckbusch et al., 2004; this study) or to most of the macroin-
vertebrates (Hsieh et al., 2002; Bouillon et al., 2002b; Guest & Connolly, 2004;
Kieckbusch et al., 2004; Abed-Navandi & Dworschak, 2005) so as to function
as a (direct and significant) source of carbon for these fauna. The most
depleted fish species in this study was M. argenteus with a mean d13C of
À24Á0%, which is far more enriched as compared to mangrove leaves. Simi-
larly, Guest & Connolly (2004) in Moreton Bay (Australia), Macia (2004) in
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1656 B. R. LUGENDO ET AL.
Inhaca Island (Mozambique) and Abed-Navandi & Dworschak (2005) on the
Belize Barrier Reef (Caribbean Sea) observed that the d13C of most crabs
and shrimps from the mangrove habitats was close to that of microphytoben-
thos and distinct from that of mangrove leaves.
Seagrasses (with exception of Halodule wrightii in the mangrove creeks and
channel with a mean d13C of À26Á3 and À20Á2%, respectively) were too far en-
riched in d13C (À15Á5 to À8Á2%) as compared to the herbivore S. sutor. This
suggests that seagrasses did not contribute to the diet of this herbivorous fish
species. The low contribution of seagrasses and the high contribution of algae
to the food web that was observed by Moncreiff & Sullivan (2001) in the Gulf
of Mexico and by Kieckbusch et al. (2004) in Biscayne Bay is another example
that seagrass plays a minor role in the food web and that algae are the primary
source of organic matter for higher trophic levels. Mangroves and seagrasses
do not appear to be direct sources of carbon in the diets of the fish species
studied; they probably serve as refugia as well as a substratum for a variety
of primary producers and consumers that are important in the food webs of
these habitats (Kieckbusch et al., 2004). Presence of food in addition to struc-
tural complexity has been reported to account for the strong association of
large numbers of juvenile fishes within mangrove forests (Laegdsgaard & Johnson,
2001). In addition, seagrass beds have also been reported to harbour a high
abundance of small invertebrates that are an important food of many juvenile
fish species (Nakamura & Sano, 2005).
Using stable carbon isotope analysis different habitats were distinguished,
which functioned as a source of carbon. Fish species from the mangroves were
more depleted in d13C as compared to individuals of the same species caught
from either the mud and sand flats or seagrass habitats. Similarly, fish species
from the mud and sand flats were more depleted relative to individuals of the
same species occurring in seagrass beds. The d13C of food also showed this
trend. This is in agreement with other studies showing that the importance
of mangrove-derived carbon (if any) is limited to the surroundings of the man-
grove habitats, and decreases when moving away from the mangroves (Rodelli
et al., 1984; Newell et al., 1995; Dehairs et al., 2000; Chong et al., 2001; Guest
& Connolly, 2004). In agreement with Dehairs et al. (2000), this observation
calls for critical evaluation on the assumption that mangrove ecosystem repre-
sent a source of organic nutrients for the coastal ecosystems. Like in other
studies from around the world, the present study shows significant feeding of
fishes (and macrobenthos) in the mangroves (Rodelli et al., 1984; Marguillier
`
et al., 1997; Sheaves & Molony, 2000; Chong et al., 2001; Cocheret de la Moriniere
et al., 2003; Guest & Connolly, 2004; Nagelkerken & van der Velde, 2004;
Abed-Navandi & Dworschak, 2005).
The overlap in stable carbon isotopes of some fish species in different bay
habitats suggests connectivity between these habitats, with the possibility that
fishes used more than one habitat as a feeding ground. Some fish species (G.
filamentosus, L. lentjan and S. sutor) from the mud and sand flats showed a pos-
sible connection to the mangrove habitats as feeding habitats. Likewise, some
fish species (G. oyena, L. fulviflamma, P. quadrilineatus and S. sutor) from Chwaka
seagrass beds showed some evidence of using mud and sand flats as feeding
habitats. An explanation for this observation could firstly be recent ontogenetic
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1657
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
`
migration (Cocheret de la Moriniere et al., 2003). The fishes could have
migrated from one habitat to another habitat recently, as a result of which they
still show part of the signature of their previously used habitat. It could take
several weeks to months to acquire the signature of the food from the new hab-
itat (Gearing, 1991; Hobson, 1999; Nagelkerken & van der Velde, 2004).
Since the fish samples were collected during low tide, a second possibility is
that fishes migrated with the tides (with a spring tidal difference of 2 m) from
the mangroves to the mud and sand flats and from the mud and sand flats to
Chwaka seagrass beds. Migration in relation to feeding (Reis & Dean, 1981),
preference of particular salinities (Quinn & Kojis, 1987) and avoidance of being
stranded during low tide in areas that fall dry (van der Veer & Bergman, 1986)
has been suggested to be among the reasons that can trigger tidal migrations.
Since the d13C signature showed intermediate values between habitats, this
could suggest that they fed at low tide as well as high tide, in two different hab-
itats. Tidal migration between bay habitats in Chwaka Bay by Lutjanidae has
been shown by Dorenbosch et al. (2004), and other species possibly follow the
same pattern of behaviour.
Distance to be covered during (tidal) migration, however, seems important in
terms of energy budget especially when juvenile fishes (<20 cm length) are con-
sidered, in which case long-distance migration costs may exceed energy intake
(Nøttestad et al., 1999). This may also be the case in the present study (in
which the majority of the fishes were 5–10 cm LF) where there appears to be
substantial connectivity for fish species between neighbouring habitats, but
not between habitats that were located far away from one another, such as
Marumbi seagrass beds located 8 and 6 km away from the mangrove and
mud and sand flat habitats, respectively.
The significant difference observed for some species in stable carbon isotopes
in individuals of the same species and similar size classes between bay habitats
suggests two situations: 1) the individuals of each habitat belong to different
assemblages, each depending completely (in terms of nutrition) on different
bay habitats, and 2) the different bay habitats all have the potential of provid-
ing sufficient food sources to the fish assemblage found therein. The differences
in fish densities of particular species and size class in different bay habitats as
observed by Lugendo et al. (2005), however, suggests that other factors than
food alone control the distribution of juvenile fishes. As observed from other
studies, structural complexity and shade in relation to predation risk are
among the important factors in determining distribution of juvenile fishes
`
(Laegdsgaard & Johnson, 2001; Cocheret de la Moriniere et al., 2004; Verweij
et al., 2006).
In conclusion, this study revealed that significant differences in stable isotope
signatures (C and N) exist in food and fishes from different bay habitats in
Chwaka Bay, which could be used to delineate feeding habitats of fishes. Fishes
appear to forage in all studied bay habitats. Seagrasses and mangroves do not
appear to be direct sources of carbon in the diets of studied fish species; rather,
they probably serve as refuge as well as a substratum for a variety of primary
producers and consumers that are important in the food webs of these habitats.
Some fish species of similar feeding guilds showed some degree of segregation
by feeding on different food resources. Zoobenthivores, however, showed an
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1658 B. R. LUGENDO ET AL.
overlap in diet and mainly fed on copepods, shrimps and crabs. There appears
to exist a connectivity for some fish species between different bay habitats with
respect to feeding (between the mud and sand flats and the mangroves, and
between the seagrass beds and the mud and sand flats), which could be a result
of either ontogenetic or tidal migration.
Gratitude is expressed to M. A. Manzi, S. J. Simgeni and M. A. Makame for assis-
tance with the fieldwork. The administration and staff of the Institute of Marine Sciences
in Zanzibar provided logistical support and research facilities. This study was financially
supported by NUFFIC through the ENVIRONS-MHO Project implemented by the Fac-
ulty of Science, University of Dar es Salaam, Tanzania, Schure-Beijerinck-Popping Foun-
dation and Quo Vadis Fonds (Radboud University), The Netherlands. I. N. was
supported by a Vidi grant from the Netherlands Organisation for Scientific Research
(NWO). This is Centre for Wetland Ecology publication no. 415.
References
Abed-Navandi, D. & Dworschak, P. C. (2005). Food sources of tropical thalassidean
shrimps: a stable-isotope study. Marine Ecology Progress Series 291, 159–168.
Beck, M. W., Heck, K. L. Jr, Able, K. W., Childers, D. L., Eggleston, D. B., Gillanders,
B. M., Halpern, B., Hays, C. G., Hoshino, K., Minello, T. J., Orth, R. J., Sheridan,
P. F. & Weinstein, M. P. (2001). The identification, conservation, and management
of estuarine and marine nurseries for fish and invertebrates. BioScience 51,
633–641.
Bouillon, S., Koedam, N. & Dehairs, F. (2002a). Primary producers sustaining macro-
invertebrate communities in intertidal mangrove forests. Oecologia 130, 441–448.
Bouillon, S., Raman, A. V., Dauby, P. & Dehairs, F. (2002b). Carbon and nitrogen stable
isotope ratios of subtidal benthic invertebrates in an estuarine mangrove ecosystem
(Andhra Pradesh, India). Estuarine Coastal and Shelf Science 54, 901–913.
Cederlof, U., Rydberg, L., Mgendi, M. & Mwaipopo, O. (1995). Tidal exchange in
¨
a warm tropical lagoon: Chwaka Bay, Zanzibar. Ambio 24, 458–464.
Chittaro, P. M., Fryer, B. J. & Sale, R. (2004). Discrimination of French grunts
(Haemulon flavolineatum Desmarest, 1823) from mangrove and coral reef habitats
using otolith microchemistry. Journal of Experimental Marine Biology and Ecology
308, 168–183.
Chong, V. C., Sasekumar, A., Leh, M. U. C. & Cruz, R. D. (1990). The fish and prawn
communities of a Malaysian coastal mangrove ecosystem, with comparisons to
adjacent mud/sand flats and inshore waters. Estuarine Coastal and Shelf Science
31, 703–722.
Chong, V. C., Low, C. B. & Ichikawa, T. (2001). Contribution of mangrove detritus to
juvenile prawn nutrition: a dual stable isotope study in a Malaysian mangrove
forest. Marine Biology 138, 77–86.
`
Cocheret de la Moriniere, E., Pollux, B. J. A., Nagelkerken, I., Hemminga, M. A.,
Huiskes, A. H. L. & van der Velde, G. (2003). Ontogenetic dietary changes of coral
reef fishes in the mangrove-seagrass-reef continuum: stable isotope and gut-content
analysis. Marine Ecology Progress Series 246, 279–289.
`
Cocheret de la Moriniere, E., Nagelkerken, I., van der Meij, H. & van der Velde, G.
(2004). What attracts juvenile coral reef fish to mangroves: habitat complexity or
shade? Marine Biology 144, 139–145.
Dehairs, F., Rao, R. G., Chandra Mohan, P., Raman, A. V., Marguillier, S. & Hellings,
L. (2000). Tracing mangrove carbon in suspended matter and aquatic fauna of the
Gautami–Godavari Delta, Bay of Bengal (India). Hydrobiologia 431, 225–241.
DeNiro, M. J. & Epstein, S. (1978). Influence of diet on the distribution of carbon
isotopes in animals. Geochimica et Cosmochimica Acta 45, 341–351.
Dorenbosch, M., Verweij, M. C., Nagelkerken, I., Jiddawi, N. & van der Velde, G.
(2004). Homing and daytime tidal movements of juvenile snappers (Lutjanidae)
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1659
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
between shallow-water nursery habitats in Zanzibar, western Indian Ocean.
Environmental Biology of Fishes 70, 203–209.
Field, A. (2000). Discovering Statistics Using SPSS for Windows. London: Sage Publica-
tions.
Fry, B. & Sherr, E. B. (1984). d13C measurements as indicators of carbon flow on marine
and freshwater ecosystems. Contributions in Marine Science 27, 13–47.
Gearing, J. N. (1991). The study of diet and trophic relationships through natural
abundance 13C. In Carbon Isotope Techniques (Coleman, D. C. & Fry, B., eds), pp.
201–218. San Diego, CA: Academic Press.
Guest, M. A. & Connolly, R. M. (2004). Fine-scale movement and assimilation of
carbon in saltmarsh and mangrove habitats by resident animals. Aquatic Ecology
38, 599–609.
´
Hartill, B. W., Morrison, M. A., Smith, M. D., Boubee, J. & Parsons, D. M. (2003).
Diurnal and tidal movements of snapper (Pagrus auratus, Sparidae) in an estuarine
environment. Marine & Freshwater Research 54, 931–940.
Hobson, K. A. (1999). Tracing origins and migration of wildlife using stable isotopes:
a review. Oecologia 120, 314–326.
Hsieh, H., Chen, C., Chen, Y. & Yang, H. (2002). Diversity of benthic organic matter
flows through polychaetes and crabs in a mangrove estuary: d13C and d34S signals.
Marine Ecology Progress Series 227, 145–155.
Hyslop, E. J. (1980). Stomach content analysis – a review of methods and their
application. Journal of Fish Biology 17, 411–429.
Khalaf, M. A. & Kochzius, M. (2002). Changes in trophic community structure of shore
fishes at an industrial site in the Gulf of Aqaba, Red Sea. Marine Ecology Progress
Series 239, 287–299.
Kieckbusch, D. K., Koch, M. S., Serafy, J. E. & Anderson, W. T. (2004). Trophic
linkages among primary producers and consumers in fringing mangroves of
subtropical lagoons. Bulletin of Marine Science 74, 271–285.
Laegdsgaard, P. & Johnson, C. R. (2001). Why do juvenile fish utilise mangrove habitats?
Journal of Experimental Marine Biology and Ecology 257, 229–253.
Little, M. C., Reay, P. J. & Grove, S. J. (1988). The fish community of an East African
mangrove creek. Journal of Fish Biology 32, 729–747.
Lugendo, B. R., Pronker, A., Cornelissen, I., de Groene, A., Nagelkerken, I.,
Dorenbosch, M., van der Velde, G. & Mgaya, Y. D. (2005). Habitat utilisation
by juveniles of commercially important fish species in a marine embayment in
Zanzibar, Tanzania. Aquatic Living Resources 18, 149–158.
Lugendo, B. R., de Groene, A., Cornelissen, I., Pronker, A., Nagelkerken, I., van der
Velde, G. & Mgaya, Y. D. (in press). Spatial and temporal variation in
fish community structure of a marine embayment in Zanzibar, Tanzania.
Hydrobiologia.
MacDonald, J. S., Waiwood, K. G. & Green, R. H. (1982). Rates of different digestion
of different prey in Atlantic cod (Gadus morhua), ocean pout (Macrozoarces
americanus), winter flounder (Pseudopleuronectes americanus), and American plaice
(Hippoglossoides platessoides). Canadian Journal of Fisheries and Aquatic Sciences
39, 651–659.
Macia, A. (2004). Primary carbon sources for juvenile penaeid shrimps in a mangrove-
fringed bay of Inhaca Island, Mozambique: a dual carbon and nitrogen isotope
analysis. Western Indian Ocean Journal of Marine Science 3, 151–161.
Mahongo, S. B. (1997). Surface heat exchange and bottom reflection in a tidal, shallow
tropical lagoon: Chwaka Bay, Zanzibar. MSc Thesis, Gothenburg University,
Sweden.
Marguillier, S., van der Velde, G., Dehairs, F., Hemminga, M. A. & Rajagopal, S. (1997).
Trophic relationships in an interlinked mangrove-seagrass ecosystem as traced by
d13C an d15N. Marine Ecology Progress Series 151, 115–121.
Minagawa, W. & Wada, E. (1984). Stepwise enrichment of 15N along food chains: further
evidence and the relation between d15N and animal age. Geochimica et Cosmochi-
mica Acta 48, 1135 –1140.
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
1660 B. R. LUGENDO ET AL.
Mohammed, S. M., Johnstone, R. W., Widen, B. & Jordelius, E. (2001). The role of
mangroves in the nutrient cycling and productivity of adjacent seagrass commu-
nities, Chwaka Bay, Zanzibar. In Marine Science Development in Tanzania and
Eastern Africa. Proceedings of the 20th Anniversary Conference on Advances in
Marine Science in Tanzania, 1999 (Richmond, M. & Francis, J., eds), pp. 205–226.
Zanzibar: IMS/WIOMSA.
Moncreiff, C. A. & Sullivan, M. J. (2001). Trophic importance of epiphytic algae in
subtropical seagrass beds: evidence from multiple stable isotope analyses. Marine
Ecology Progress Series 215, 93–106.
Nagelkerken, I. & van der Velde, G. (2004). Relative importance of interlinked
mangroves and seagrass beds as feeding habitats for juvenile reef fish on a
Caribbean island. Marine Ecology Progress Series 274, 153–159.
`
Nagelkerken, I., Dorenbosch, M., Verberk, W. C. E. P., Cocheret de la Moriniere, E. &
van der Velde, G. (2000). Day-night shifts 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.
Nakamura, Y. & Sano, M. (2005). Comparison of invertebrate abundance in a seagrass
bed and adjacent coral and sand areas at Amitori Bay, Iriomote Island, Japan.
Fisheries Science 71, 573–550.
Newell, R. I. E., Marshall, N., Sasekumar, A. & Chong, V. C. (1995). Relative
importance of benthic microalgae, phytoplankton, and mangroves as sources of
nutrition for penaeid prawns and other coastal invertebrates from Malaysia.
Marine Biology 123, 595–606.
Nøttestad, L., Giske, J., Chr. Holst, J. & Huse, G. (1999). A length-based hypothesis for
feeding migrations in pelagic fish. Canadian Journal of Fisheries and Aquatic
Sciences 56, 26–34.
Parrish, J. D. (1989). Fish communities of interacting shallow-water habitats in tropical
oceanic regions. Marine Ecology Progress Series 58, 143–160.
Pinnegar, J. K. & Polunin, N. V. C. (1999). Differential fractionation of d13C and d15N
among fish tissues: implication for the study of trophic interactions. Functional
Ecology 13, 225–231.
Polis, G. & Strong, D. R. (1996). Food web complexity and community dynamics.
American Naturalist 147, 813–846.
Quinn, N. J. & Kojis, B. L. (1987). The influence of diel cycle, tidal direction and trawl
alignment on beam trawl catches in an equatorial estuary. Environmental Biology of
Fishes 19, 297–308.
Rao, R. G., Woitchik, A. F., Goeyens, L., van Riet, A., Kazungu, J. & Dehairs, F.
(1994). Carbon, nitrogen contents and stable isotope abundance in mangrove
leaves from an east African coastal lagoon (Kenya). Aquatic Botany 47, 175–183.
Reis, R. R. & Dean, J. M. (1981). Temporal variations in the utilization of an intertidal
creek by the Bay Anchovy (Anchoa mitchilli). Estuaries 4, 16–23.
Robertson, A. I. & Blaber, S. J. M. (1992). Plankton, epibenthos and fish communities.
In Tropical Mangrove Ecosystems (Robertson, A. I. & Alongi, D. M., eds), pp.
173–224. Washington, DC: American Geographical Union.
Robertson, A. I. & Duke, N. C. (1987). Mangrove as nursery sites: comparison of the
abundance and species composition of fish and crustaceans in mangroves and
other near shore habitats in tropical Australia. Marine Biology 96, 193–205.
Rodelli, M. R., Gearing, J. N., Gearing, P. J., Marshall, N. & Sasekumar, A. (1984).
Stable isotope ratios as a tracer of mangrove carbon in Malaysian ecosystems.
Oecologia 61, 326–333.
Rooker, J. R. & Dennis, G. D. (1991). Diel, lunar and seasonal changes in a mangrove
fish assemblage off southwestern Puerto Rico. Bulletin of Marine Science 49,
684–698.
Sheaves, M. & Molony, B. (2000). Short-circuit in the mangrove food chain. Marine
Ecology Progress Series 199, 97–109.
Smith, M. M. & Heemstra, P. C. (Eds) (1991). Smith’s Sea Fishes, 1st edn. Johannesburg:
Southern Book Publishers.
# 2006 The Authors
Journal compilation 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661
#
1661
FEEDING GROUNDS FOR TROPICAL JUVENILE FISHES
Sydeman, W. J., Hobson, K. A., Pyle, P. & McLaren, E. B. (1997). Trophic relationships
among seabirds in Central California. The Condor 99, 327–336.
de Troch, M., Mees, J. & Wakwabi, E. (1998). Diets of abundant fishes from beach seine
catches in seagrass beds of a tropical bay (Gazi Bay, Kenya). Belgian Journal of
Zoology 128, 135–154.
Vance, D. J., Haywood, M. D. E., Heales, D. S., Kenyon, R. A., Loneragan, N. R. &
Pendrey, R. C. (1996). How far do prawns and fish move into mangroves?
Distribution of juvenile banana prawns, Penaeus merguiensis, and fish in a tropical
mangrove forest in northern Australia. Marine Ecology Progress Series 131,
115–124.
Vanderklift, M. A. & Ponsard, S. (2003). Sources of variation in consumer-diet d15N
enrichment: a meta-analysis. Oecologia 136, 169–182.
van der Veer, H. M. & Bergman, M. N. J. (1986). Development of tidally related
behaviour of a newly settled 0-group plaice (Pleuronectes platessa) populations in
the western Wadden Sea. Marine Ecology Progress Series 31, 121–129.
Verweij, M. C., Nagelkerken, I., de Graaff, D., Peeters, M., Bakker, E. J. & van der
Velde, G. (2006). Structure, food and shade attract juvenile coral reef fish to
mangrove and seagrass habitats: a field experiment. Marine Ecology Progress Series
306, 257–268.
Electronic Reference
Froese, R. & Pauly, D. (Eds) (2004). FishBase. Available at www.fishbase.org (version
08/2004).
# 2006 The Authors
Journal compilation # 2006 The Fisheries Society of the British Isles, Journal of Fish Biology 2006, 69, 1639–1661