Dual isotope study of food sources of a fish assemblage in the red river mangrove ecosystem, vietnam

The results suggested that mangrove carbon contributed a small proportion in the diets of the mangrove fish, with dominant food sources coming from benthic inverte-brates, including ocyp

B I O D I V E R S I T Y I N A S I A N C O A S T A L W A T E R S

Dual isotope study of food sources of a fish assemblage

in the Red River mangrove ecosystem, Vietnam

Nguyen Tai Tue •Hideki Hamaoka•Tran Dang Quy•

Mai Trong Nhuan•Atsushi Sogabe•Nguyen Thanh Nam•

Koji Omori

Received: 24 February 2013 / Accepted: 27 October 2013

Ó Springer Science+Business Media Dordrecht 2013

Abstract The food source utilization and trophic

relationship of the fish assemblage in the Red River

mangrove ecosystem, Vietnam were examined using

dual isotope analysis The carbon and nitrogen stable

isotope signatures of 23 fish species ranged from

-24.0 to -15.7% and from 8.8 to 15.5%, respectively

Cluster analysis based on the d13C and d15N signatures

clearly separated the mangrove fish into five feeding

groups, representing detritivores, omnivores,

pisci-vores, zoobenthipisci-vores, and zooplanktipisci-vores, which

concurred with the dietary information The results suggested that mangrove carbon contributed a small proportion in the diets of the mangrove fish, with dominant food sources coming from benthic inverte-brates, including ocypodid and grapsid crabs, penaeid shrimps, bivalves, gastropods, and polychaetes The

d15N values showed that the food web structure may be divided into different trophic levels (TLs) The lowest TLs associated with Liza macrolepis, Mugil cephalus, and Periophthalmus modestus; 18 fish species had TLs between 3.0 and 3.8; and Pennahia argentata had the highest TL (c 4.0)

Keywords Mangrove ecosystem Stable isotopes Fish  Food sources  Trophic level  Vietnam

Introduction

Mangrove ecosystems have often been considered as hot spots of fish diversity (Nagelkerken et al.,2008) The hypotheses of the high diversity of the fish in the mangrove ecosystem include reduced predation, increased living space, and dominated food supply (Nagelkerken et al.,2008) The last of these hypoth-eses stated that the mangrove ecosystem produces greater food densities such as mangrove detritus, benthic microalgae (BMA), sediment organic matter, infauna, and invertebrates that form the basal food sources of the fish in the mangrove ecosystem

Guest editors: M Tokeshi & H T Yap / Biodiversity in

Changing Coastal Waters of Tropical and Subtropical Asia

N T Tue

Graduate School of Science and Engineering, Ehime

University, 2-5 Bunkyo-cho, Matsuyama, Japan

N T Tue ( &)  H Hamaoka  K Omori

Center for Marine Environmental Studies, Ehime

University, 2-5 Bunkyo-cho, Matsuyama 790-8577, Japan

e-mail: tuenguyentai@gmail.com

T D Quy  M T Nhuan

Faculty of Geology, VNU University of Science, 334

Nguyen Trai, Thanh Xuan, Hanoi, Vietnam

A Sogabe

Research Center for Marine Biology, Asamushi, Tohoku

University, 9 Sakamoto, Asamushi, Aomori 039-3501,

Japan

N T Nam

Faculty of Biology, VNU University of Science, 334

Nguyen Trai, Thanh Xuan, Hanoi, Vietnam

DOI 10.1007/s10750-013-1737-9

The linkage between the mangroves and the fish has

been the focus of numerous studies (e.g., Odum &

Heald,1972; Blaber,2007; Layman,2007;

Nagelker-ken et al., 2008) Based on the stomach content

analysis, Odum & Heald (1972) demonstrated that the

mangrove detritus was a predominant food source of

the fish in the estuarine habitats Nevertheless, isotopic

studies have failed to confirm the contribution of the

mangrove detritus in the diets of the fish in the

estuarine food web (Rodelli et al.,1984), apparently

because the refractory mangrove detritus is not easily

digested by the gut system of the fish (Fry & Ewel,

2003) Subsequently, numerous studies have focused

on determining the functional relationship between the

mangroves and the fish through examining the food

source utilization (Sheaves & Molony, 2000; Nanjo

et al.,2008), and the trophic relationship (Abrantes &

Sheaves, 2009; Giarrizzo et al., 2011) The results

would be useful for ecosystem-based fishery

manage-ment and mangrove conservation practices

(Nagel-kerken et al.,2008)

The stable isotope ratios of carbon (d13C) and of

nitrogen (d15N) are useful in determining the

time-averaged relative importance of the ingested food

sources and the relative trophic level (TL) of a

consumer (Michener & Lajtha, 2007) The mean

(±1SD) trophic enrichment factor (TEF) between an

animal and its diet for d13C and d15N is 0.4 ± 1.3 and

3.4 ± 1%, respectively (Post, 2002) Therefore, the

d13C values can be used to trace the carbon utilization

by an organism when the stable isotope signatures of

the food sources are different (Bouillon et al.,2008)

In addition, the d15N values can be used to estimate the

relative TL of the organism (Zanden & Rasmussen,

1999; Post,2002)

The d13C and d15N values have been frequently used

to examine the food sources (Sheaves & Molony,

2000) and the TL of the fish (Abrantes & Sheaves,

2009) in the mangrove ecosystem, and the ecological

connectivity between mangrove forests and other

coastal ecosystems (Layman, 2007) The isotopic

studies have shown that the contribution of the

mangrove carbon in the diets of the fish varies by

landscape characteristics (Thimdee et al.,2004;

Lug-endo et al.,2007) and tidal water levels (Sheaves &

Molony,2000) of the mangrove ecosystem However,

the application of the stable isotope methods in

understanding ecological functions of the mangrove

ecosystem in Vietnam has been few, and there remains

a significant gap in knowledge concerning the impor-tance of horizontal and vertical trophic dynamics of the fish assemblage within and between adjacent systems These problems certainly constrain our understanding

of the importance of the mangroves to fisheries, the valuation of ecological services of mangroves, and the planning and implementation effective conservation

In the present study, we analyzed the stable isotope signatures of carbon (d13C) and nitrogen (d15N) of 183 individuals from 23 fish species, and used the isotopic data of primary production (mangroves, BMA, and marine phytoplankton), mangrove creek particulate organic matter (POM), sediments, and major groups of the benthic invertebrates for testing the hypothesis of whether the mangrove carbon was a major food source

of the fish assemblage in the Red River mangrove ecosystem of Vietnam To test our hypothesis, two objectives were investigated: (1) to determine the utilization of food sources by the fish assemblage and (2) to determine the relative trophic relationship of the fish assemblage in the Red River mangrove ecosystem

of Vietnam

Materials and methods

Study area

The present study was conducted in the Red River Delta Biosphere Reserve (RRBR) in northern Viet-nam The RRBR has two primary mangrove forests, the Xuan Thuy National Park and the Tien Hai Natural Reserve (Fig.1) (http://www.unesco.org) The char-acteristics of the mangrove forests are earlier descri-bed in Tue et al (2011, 2012a, c) Briefly, the mangrove forests are predominated by Sonneratia caseolaris (L.) Engl., Bruguiera gymnorrhiza (L.) Lamk., Kandelia candel (L.) Druce, and Aegiceras corniculatum (L.) Blanco, and consist of several major creek systems (Fig.1), which remain inundated throughout the tidal regimes Moreover, mangroves are an important source of the POM (Tue et al.,

2012b), and important sinks to organic carbon and fine sediment particles (Tue et al.,2012d) As a result, the mangrove forests are thought to provide productive food sources for the benthic invertebrates (Tue et al.,

2012c) and the fish in the estuarine habitats (Cuong & Khoa,2004) The mangrove forests also play impor-tant roles in the filtering and containment of various

pollutants (i.e., trace elements; Tue et al., 2012a), as

well as a physical buffer against erosion and surge

from major storm events Moreover, the mangrove

forests are of great importance as major feeding,

breeding, and stopover grounds for migratory birds,

including several highly threatened species Platalea

minor (Temminck & Schlegel, 1849), Larus

ichthya-etus (Pallas, 1773), Tringa orchropus (Linnaeus,

1758), and Egretta eulophotes (Swinhoe, 1860)

(Nhuan et al.,2009)

Field sampling

Fish samples were collected by a gill net during spring

and ebb tides in two major tidal creeks of the Xuan

Thuy National Park and the Tien Hai Nature Reserve

(Fig.1) in January–February 2008 The sampling was

designed to collect most predominant fish species and

those of high economic values (Cuong & Khoa,2004;

Than, 2004) Fish samples were placed in labeled polyethylene bags, immediately stored in ice, and transported to the laboratory where they were frozen at -20°C until processing and analysis

Sample preparation and analysis

In the laboratory, the fish samples were first rinsed with distilled water, wiped with paper towel, then identified to species level, measured for total body length, and categorized into feeding groups based on the literature (Balan, 1967; Koslow, 1981; Elliott

et al., 2007; Platell et al., 2007; Baeck et al.,2008; Nanjo et al., 2008; Salameh et al., 2010; Froese & Pauly,2011) The list of the mangrove fish observed in the RRBR and their feeding ecology is shown in Table 1 These feeding groups consisted of detriti-vores (detritus and/or microphytobenthos feeders), omnivores (filamentous algae, macrophytes,

Fig 1 Sampling sites

within the Red River

mangrove ecosystem,

Vietnam

Table 1 The list of fishes observed in the Red River mangrove ecosystem, Vietnam

ecology

References Taxa

Anguilliformes

Aulopiformes

Harpadon nehereus (Hamilton, 1822) Nekton and fishes PV Froese & Pauly ( 2011 ) Muraenesox cinereus (Forsska˚l, 1775) Fishes and crustaceans PV Froese & Pauly ( 2011 ) Clupeiformes

Escualosa thoracata (Valenciennes, 1847) Zooplankton (copepods, crab zoea,

larvae of bivalves, and fish eggs) and phytoplankton

ZP Froese & Pauly ( 2011 )

Coilia mystus (Linnaeus, 1758) Zooplankton and phytoplankton ZP Koslow ( 1981 ) Perciformes

Acanthopagrus latus (Houttuyn, 1782) Mangrove detritus, sesarmid crabs,

small gastropods, worms, crustaceans, and mollusks

ZB Platell et al ( 2007 )

Bostrychus sinensis (Lacepe`de, 1801) Crustaceans and small fishes PV Froese & Pauly ( 2011 ) Butis butis (Hamilton, 1822) Small fishes and crustaceans PV Froese & Pauly ( 2011 ) Gerres limbatus (Cuvier, 1830) Small benthic animals ZB Froese & Pauly ( 2011 ) Glossogobius biocellatus

(Valenciennes, 1837)

Fishes, detritus, and gammaridean amphipods

PV Nanjo et al ( 2008 )

Leiognathus bindus (Valenciennes, 1835) Copepods, phytoplankton, detritus,

and zoobenthos

ZP Balan ( 1967 ) Liza macrolepis (Smith, 1846) Algae, diatoms, forams, benthic

polychaete, crustaceans, mollusks, organic matter, and detritus;

copepods and floating algae

DV Froese & Pauly ( 2011 )

Lutjanus russellii (Bleeker, 1849) Crabs, shrimps, fishes, crustaceans,

and insects

PV Nanjo et al ( 2008 ) Mugil cephalus (Linnaeus, 1758) Zooplankton as larvae; detritus,

microalgae, and benthic organisms

DV Nanjo et al ( 2008 ) Oxyeleotris marmorata (Bleeker, 1852) Small fishes, shrimps, aquatic insects,

mollusks, and crabs

PV Froese & Pauly ( 2011 )

Pennahia argentata (Houttuyn, 1782) Small fishes and invertebrates PV Froese & Pauly ( 2011 ) Periophthalmus modestus (Cantor, 1842) Gammarids, crabs, and other

crustaceans

OV Baeck et al ( 2008 )

Sillago sihama (Forsska˚l, 1775) Polychaete worms, small prawns

(penaeus), shrimps, and amphipods

ZB Froese & Pauly ( 2011 ) Terapon theraps (Cuvier, 1829) Animals ZB Froese & Pauly ( 2011 ) Trypauchen vagina (Bloch & Schneider, 1801) Small invertebrates and crustaceans ZB Salameh et al ( 2010 ) Scorpaeniformes

General feeding ecology and reference are shown for each fish species

NA Not available, DV detritivores, OV omnivores, PV piscivores, ZB zoobenthivores, and ZP zooplanktivores

periphyton, epifauna and infauna feeders),

zooplankti-vores (zooplankton, hydroids, planktonic crustacean,

and fish eggs/larval feeders), zoobenthivores (benthic

invertebrate feeders), and piscivores (finfish and

nektonic invertebrate feeders) (Elliott et al.,2007)

The processing of fish for the stable isotope analysis

involved the extraction of white muscle tissue from the

anterior dorsal region The white tissue is more

isotopically homogenous than other tissues (Michener

& Lajtha,2007) The fish tissues were then placed in

Eppendorf tubes, dried in an electric oven at 60°C for

24 h, and ground to fine powder by an agate mortar

and pestle The lipids were extracted from the fish

tissues prior to the stable isotope analysis following

methods described in Tue et al (2012c) Briefly, the

pulverized fish tissues were placed in the Eppendorf

tubes, immersed in a 2:1 chloroform:methanol (by

volume) solution, and left at room temperature for

24 h to extract the lipids The samples were then

rinsed with distilled water, and dried in an electric

oven at 60°C for 24 h

For all samples, 1.0 ± 0.1 mg of the pulverized fish

tissues was packed in a tin capsule The carbon and

nitrogen stable isotope signatures were measured

using an isotope ratio mass spectrometer

(ANCA-GSL; Sercon Inc, UK) and expressed in d notion as

parts per thousand (permil, %) as shown in Eq (1):

d13C or d15N ¼ Rsample

Rstd

 1

 1000 ð1Þ where R is isotope ratios13C/12C or15N/14N Rsampleis

the isotope ratio of the sample, and Rstdis the isotope

ratio of a standard referenced to Pee Dee Belemnite

limestone carbonate (PDB) for d13C, and to

atmo-spheric nitrogen for d15N During analysis processes,

L-histidine (d13C = -11.4% and d15N = -7.6%)

was used for quantifying the analyzed results

Ana-lytical errors were 0.1% for d13C and 0.2% for d15N,

respectively

Background data for potential food sources

of the mangrove fish

The ranges of the d13C and d15N values of the

mangrove leaves, the marine phytoplankton, the

BMA, the POM, the sediments, and the benthic

invertebrates are shown in Fig.2 The respective

means of the d13C values increased in the order of the

mangrove leaves, mangrove sediments, the tidal flat,

the creek bank and creek bottom sediments, the POM, the marine phytoplankton, and the BMA (Tue et al.,

2012b,c,d)

The benthic invertebrates (e.g., grapsid crabs) have been shown to be important food sources for the mangrove fish (Sheaves & Molony,2000) The ranges

of the d13C and d15N values of the major benthic invertebrate groups in the mangrove ecosystem of the RRBR are shown in Fig 2 Tue et al (2012c) reported that the gastropods, bivalves, grapsid crabs, and polychaetes inhabiting the mangrove forests directly relied on the mangrove detritus The ocypodid crabs inhabiting the land–water ecotone showed preference for the BMA and other food sources (i.e., bacteria, ciliate protozoa, and nematodes) over the mangrove detritus The diets of tidal flat bivalve Ensis magnus (Ensis Schumacher, 1817) was a mixture of the marine phytoplankton and the BMA The penaeid prawns

Fig 2 Dual isotope plot of mean d13C and d15N signatures (±1 SD) of the different food sources, and the fish in the Red River mangrove ecosystem, Vietnam Acronyms of fish taxa are shown in Table 2 , and DV, OV, PV, ZB, and ZP denotes the detritivores, omnivores, piscivores, zoobenthivores, and zoo-planktivores, respectively Mang, Mang sed, Adj sed, POM, Phyto, and BMA denotes mangrove leaves, mangrove sedi-ments, adjacent habitat sedisedi-ments, creek particulate organic matter, marine phytoplankton, and benthic microalgae, respec-tively The stable isotope data of mangrove leaves, POM, and phytoplankton; benthic microalgae and invertebrates; and mangrove, creek bank, tidal flat and bottom sediments are presented in Tue et al ( 2012a , b , c ), respectively

were opportunistic omnivorous, feeding on the BMA,

the marine phytoplankton, the POM, and juvenile

invertebrates (i.e., crabs, gastropods, and bivalves),

with the latter being predominant (Tue et al.,2012c)

Estimation of the relative trophic level

and contribution of the food sources in diets

of the mangrove fish

The diversity of the food sources generates difficulties

in establishing an isotopic baseline for estimating the

relative TL of the mangrove fish (Layman, 2007)

Instead, the d15N values of the primary consumers

(invertebrates potentially eaten by fish) were used as

an index of nitrogen isotope compositions entering the

base of the food web This method reduces the

temporal and spatial variations in the d15N values of

the primary producers (Post, 2002), and provides a

more temporally integrated measurement of the

rela-tive TL of the mangrove fish (Zanden & Rasmussen,

1999) In the present study, the relative TL of the

mangrove fish was estimated from the d15Nbasevalues

of the bivalve E magnus based on the Eq (2) (Post,

2002) The d15N values of the E magnus were used as

the isotopic baseline, because this species is a true

suspension feeder, feeding on the marine

phytoplank-ton and the BMA (Tue et al.,2012c)

TL¼ ðd15Nfish  d15NbaseÞ

where TL is the relative trophic level of the mangrove

fish; d15Nfish and d15Nbase are the nitrogen stable

isotope values of the mangrove fish and the bivalve

E magnus, respectively; the mean trophic enrichment

factor between the mangrove fish and the bivalve

E magnus for d15N is 3.4; and the TL of the bivalve

E magnus is 2

The contribution of the mangroves, the marine

phytoplankton, the BMA, the POM, the sediment

organic matters, grapsid crabs, ocypodid crabs,

E magnus, gastropods, and polychaetes to the diets

of the mangrove fish was estimated using a Stable

Isotope Analysis in R (SIAR) package (Parnell et al.,

2010) on R software (R Core Team,2012) The SIAR

package is based on a Bayesian framework that can be

used to find a solution for an isotope mixing model

(Parnell et al.,2010) The SIAR package requires the

d13C and d15N values of the mangrove fish, the mean

and SD of the d13C and d15N values for the food

sources, and the mean and SD of the trophic enrich-ment factors In the present study, the mean trophic enrichment factors (±1SD) for d13C and d15N were 0.4 ± 1.3 and 3.4 ± 1.0%, respectively (Post,2002) The stable isotope values of the sediments from mangrove forests, and the adjacent habitats (tidal flats, bank and bottom creeks) were pooled, representing the sediment organic carbon source The isotope mixing model was run 5 9 106iterations with an elimination

of the first 5 9 104 The contributions of the food sources in the diets of the mangrove fish were reported

as means and lower and upper ranges (5th and 95th percentiles)

Statistical analysis

The d13C and d15N values of the mangrove fish were used to perform hierarchical cluster analysis, which can be used to categorize the mangrove fish into feeding groups based on their similarity (Mazumder

et al.,2011) and their dietary information (Table1) The distance metric was based on the Euclidean distance completed linkage method The statistical analysis was performed using the SPSS statistical software package 17 (SPSS 17.0)

Results

Stable isotope values of the mangrove fish

The d13C and d15N values of the mangrove fish ranged from -24.0 to -15.7% and from 8.8 to 15.5%, respectively (Table2; Fig.2) The lowest and highest

d13C values were expressed in Periophthalmus mode-stus and Lutjanus russellii, respectively (Fig.2) Penna-hia argentata showed the highest mean d15N values, while the lowest mean d15N values were expressed in Mugil cephalus and Liza macrolepis (Fig.2)

Feeding groups and the contribution of different food sources to diets of the mangrove fish

Five groups of the mangrove fish were categorized at a similar index level of 15 (Fig.3) Based on the dietary information (Table1), the mangrove fish were clas-sified into five feeding groups, consisting of detriti-vores, omnidetriti-vores, piscidetriti-vores, zoobenthidetriti-vores, and zooplanktivores Detritivorous fish were M cephalus

Table 2 d 13 C and d 15 N values of the fish collected from the Red River mangrove ecosystem, Vietnam

Taxa

Anguilliformes

Aulopiformes

Harpadon nehereus Hn 6 24.8 ± 3.0

(20.7 - 28.6)

-17.1 ± 0.3 13.0 ± 0.3 3.6 ± 0.1

Muraenesox cinereus Mc 3 30.9 ± 7.6

(23.5 - 41.4)

-19.6 ± 1.1 12.4 ± 0.7 3.45 ± 0.2

Clupeiformes

Escualosa thoracata Et 3 5.3 ± 0.5

(4.7 - 5.7)

-18.7 ± 0.1 13.0 ± 0.4 3.63 ± 0.13

(11.9 - 15.8)

-20.1 ± 0.3 13.4 ± 0.1 3.73 ± 0.04

Perciformes

Acanthopagrus latus Al 12 18.2 ± 3.1

(13.3 - 22.5)

-20.8 ± 1.1 12.3 ± 0.7 3.42 ± 0.2

Bostrychus sinensis Bs 6 8.7 ± 1.0

(7.7 - 10)

-20.0 ± 0.3 11.5 ± 0.2 3.17 ± 0.06

(6.7 - 8.1)

-20.4 ± 0.6 11.9 ± 0.2 3.30 ± 0.05

Gerres limbatus Gl 10 10.8 ± 1.6

(8.9 - 13.9)

-19.1 ± 0.7 13.3 ± 0.7 3.69 ± 0.21

Glossogobius biocellatus Gb 19 11.6 ± 4.6

(5.0 - 26.5)

-20.3 ± 0.8 11.3 ± 0.4 3.11 ± 0.13

Gobiomorphus sp Gob 4 23.8 ± 2.9

(20.7 - 26.6)

-19.5 ± 1.3 11.7 ± 0.7 3.23 ± 0.2

Leiognathus bindus Lb 3 8.2 ± 0.6

(7.4 - 8.6)

-19.5 ± 1.5 13.7 ± 0.1 3.81 ± 0.03

Liza macrolepis Lm 5 18.1 ± 5.5

(13.8 - 27.3)

-18.6 ± 1.2 10.6 ± 0.9 2.91 ± 0.25

Lutjanus russellii Lr 3 11 ± 0.7

(10.2 - 11.7)

-17.2 ± 1.1 12.0 ± 0.2 3.34 ± 0.05

Mugil cephalus Mcl 18 15.8 ± 3.1

(11.1 - 21.3)

-17.4 ± 0.8 10.5 ± 0.8 2.88 ± 0.24

Oxyeleotris marmorata Om 28 8.2 ± 4.8

(3.5 - 18)

-19.8 ± 1.3 11.9 ± 0.7 3.28 ± 0.21

Parapercis sp Ps 11 12.2 ± 2.6

(7.0 - 16.4)

-19.5 ± 0.8 12.4 ± 0.4 3.44 ± 0.11

Pennahia argentata Pa 8 24.5 ± 1.4

(23.4 - 26.8)

-17.3 ± 0.5 14.2 ± 0.5 3.97 ± 0.15

Periophthalmus modestus Pm 10 3.6 ± 1.7

(1.8 - 6.6)

-21.8 ± 1.7 10.7 ± 0.8 2.95 ± 0.22

(7.4 - 9.1)

-17.9 ± 0.2 13.4 ± 0.2 3.74 ± 0.05

and L macrolepis The P modestus could be

distin-guishable with other fish groups and fed on the

filamentous algae, small invertebrates, and infauna

(Table1), representing the omnivorous fishes The

piscivorous fishes were P argentata, Sillago sihama,

Harpadon nehereus, and L russellii, and may

con-sume other fish and the invertebrates (Table1) The

zoobenthivorous fishes consisted of Acanthopagrus

latus, Butis butis, Bostrychus sinensis, Gobiomorphus

sp., Glossogobius biocellatus, Moringua sp.,

Murae-nesox cinereus, Oxyeleotris marmorata, Onigocia sp.,

Parapercis sp., and Trypauchen vagina whose prey

included invertebrates (polychaetes, crabs, and

mol-lusks) and prawns (Table1) The zooplanktivorous

fishes were Escualosa thoracata, Coilia mystus,

Leiognathus bindus, Gerres limbatus, and Terapon

theraps, feeding predominantly on copepods, crab

zoera, bivalve larvae, and fish eggs (Table1)

The isotope mixing model results showed that the

mangrove carbon was a minor contributor to the diets

of the mangrove fish (Table 3) The major carbon food

sources of the mangrove fish were the benthic

invertebrates, consisting of the panaeid prawns, the

ocypodid and grapsid crabs, gastropods, the E

mag-nus, and polychaetes (Table3)

The relative trophic level of the mangrove fish

The mean relative TL (±SD) of the mangrove fish

ranged between 2.88 ± 0.24 and 3.97 ± 0.15 (Fig 2;

Table2) The relative TLs for M cephalus, L

mac-rolepis, and P modestus were below 3.0 The relative

TLs of 18 fish species ranged between 3.0 and 3.8 The

highest TL was observed from P argentata (mean

3.97 ± 0.15), and followed by L bindus (mean 3.81 ± 0.03)

Discussion

The mangrove leaves, the marine phytoplankton, the BMA, the POM, and the sediments had different d13C signatures, but all had low d15N values The d15N values of the mangrove fish concurrently increased with the TLs in the food web (Fig.2) The d13C values were, therefore, a good indicator of the origin of the

Table 2 continued

Taxa

Terapon theraps Tt 3 11.7 ± 1.0

(10.6 - 12.4)

-18.8 ± 0.4 13.2 ± 0.6 3.67 ± 0.17

Trypauchen vagina Tv 7 14.2 ± 4.0

(8.2 - 18.7)

-20.4 ± 1.3 10.9 ± 1.5 3.00 ± 0.43

Scorpaeniformes

(8.1 - 19.4)

-19.0 ± 0.8 11.6 ± 0.4 3.21 ± 0.11

ACR acronym; n is number of the samples; mean, and mean ± 1SD values are given where n = 2, and C3, respectively; L total body length (mean ± 1SD (min - max)); Trophic levels of the mangrove fishes are estimated by the d15N values

Fig 3 Results of hierarchical cluster analysis of 23 fish species based on d 13 C and d 15 N signatures DV, OV, PV, ZB, and ZP denotes the detritivores, omnivores, piscivores, zoobenthivores, and zooplanktivores, respectively

Grapsid crabs

Ocypodid crabs

Species Periophthalmus modestus

Escualosa thoracata

Leiognathus bindus

Acanthopagrus latus

Bostrychus sinensis

Glossogobius biocellatus

Gobiomorphus sp.

Muraenesox cinereus

Oxyeleotris marmorata

Trypauchen vagina

Harpadon nehereus

Pennahia argentata

food sources (Bouillon et al.,2008), while d15N values

were indicative of the relative TLs (Michener &

Lajtha, 2007) The wide variations in the d13C and

d15N values of the mangrove fish (Fig.2; Table2)

indicated that they utilized heterogeneous diets

Moreover, the fish tissues were much more enriched

in 13C composition relative to the mangrove leaves

(Fig.2), suggesting that the fish had little reliance on

the mangrove carbon sources (Table3) This pattern

was consistent with findings in the mangrove

ecosys-tems in the Tanzanian coastal waters (Lugendo et al.,

2007) and Gazi Bay, Kenya (Nyunja et al.,2009)

Among fish taxa analyzed, the detritivorous fishes

had the lowest d15N values (Fig.2) and high BMA

proportion in their diets (Table3) The results indicated

that they fed on lower trophic order sources, such as the

BMA, the sediment organic matter, and the POM

(Tables1,3; Fig.2) This finding was consistent with

the observation of Lin et al (2007), who showed that the

preferred food sources of the L macrolepis and other

detritivorous fishes were the BMA and the POM

Moreover, the d13C values of the detritivorous fish in the

present study were much higher than those of the BMA,

the POM, and the sediments (Fig.2), suggesting that

they also fed on other13C-enriched food sources such as

the benthic invertebrates The mixing model results

showed that the ocypodid crabs contributed up to 41.7%

in the diet of M cephalus (Table3) The benthic

invertebrates could be incidentally ingested while the

detritivorous fishes were feeding on the detritus, placing

them at higher TLs than the secondary consumers in the

mangrove food web (Nanjo et al.,2008)

The d13C values of the P modestus were higher

than those of polychaetes, the POM, the sediments,

and overlapped with the d13C values of E magnus and

gastropods (Fig.2) The mixing model results showed

that the major food sources of the P modestus were

polychaetes and other invertebrates (Table3) This is

consistent with Baeck et al.’s (2008) observation that

the major food items of the Periophthalmus species

were gammarid amphipods, crabs, other crustaceans,

and benthic organisms

The zooplanktivorous fishes C mystus, E

thora-cata, G limbatus, L bindus, and T theraps were

closely positioned in the food web (Fig.2) and clustered

in the same group (Fig.3), indicating that they had

similar feeding behaviors The mixing model results

showed that the major prey items of the zooplankton

fishes were the grapsid and ocypodid crabs, the penaeid

prawns, gastropods, and bivalves (Table3) The diets of the zooplanktivorous fishes in the present study were in reasonable agreement with information on their feeding ecology from the literature (Balan, 1967; Koslow,

1981) In which, the anchovy C mystus and sardine

E thoracata are suspension feeders, feeding on a diversity of available zooplankton, fish eggs, and the invertebrate larvae, rather than selecting specific species (Koslow, 1981) In addition, L bindus is reported to feed on copepods, bivalve larvae, crustaceans, and marine phytoplankton (Balan,1967)

Despite the wide feeding preferences of the zoobenthivorous fishes (Nanjo et al.,2008; Froese & Pauly, 2011), their d13C and d15N values varied slightly (Table2; Fig.2), indicating that they could feed on similar food sources, consisting of the penaeid prawns, the ocypodid and grapsid crabs, and the bivalve (E magnus) (Table3) The food sources of the zoobenthivorous fishes were consistent with the dietary information from the literature (Platell et al.,

2007; Froese & Pauly, 2011) For example, the

A latus collected from the mangrove ecosystem from Shark Bay (Australia) fed predominantly on the sesarmid crabs, small gastropods, and the mangrove materials (Platell et al., 2007) The present study showed that the d13C values of the A latus were higher than those of the bivalve E magnus, and overlapped with the d13C values of the gastropods and grapsid crabs (Fig.2) In addition, the bivalve

E magnus, ocypodid crabs, polychaetes, and prawns were predominant food items of the A latus (Table3) The low contribution of the mangrove detritus in their diets could be interpreted by either the selective feeding mechanisms or assimilation efficiency of the

A latus The A latus could predominantly ingest the benthic invertebrates selectively and reject the man-grove detritus In that case both the benthic inverte-brates and the mangrove detritus were simultaneously ingested by the fish, yet the mangrove detritus was too refractory for assimilation (Fry & Ewel,2003) The d13C values of the piscivorous fishes were overlapped, and higher than those of other fish groups from 1.3 to 4.1% (Table2; Fig.2); hence, they may not extensively feed upon other fishes The mixing model results showed that the grapsid and ocypodid crabs, and the penaeid prawns were their major food sources (Table3) Furthermore, the benthic invertebrates in the mangrove ecosystem of the RRBR are known to consume the mangrove detritus, the BMA, marine