• 沒有找到結果。

Analyses of stomach contents and stable isotopes reveal food sources of estuarine detritivorous fish in tropical/subtropical Taiwan

N/A
N/A
Protected

Academic year: 2021

Share "Analyses of stomach contents and stable isotopes reveal food sources of estuarine detritivorous fish in tropical/subtropical Taiwan"

Copied!
11
0
0

加載中.... (立即查看全文)

全文

(1)

Analyses of stomach contents and stable isotopes reveal

food sources of estuarine detritivorous fish in

tropical/subtropical Taiwan

Hsing-Juh Lin

a,b,

*

, Wen-Yuan Kao

c

, Ya-Ting Wang

a aDepartment of Life Sciences, National Chung Hsing University, Taichung 402, Taiwan, ROC

bInstitute of Marine Environmental Chemistry and Ecology, National Taiwan Ocean University, Keelung 202, Taiwan, ROC cDepartment of Life Science, National Taiwan University, Taipei 106, Taiwan, ROC

Received 3 August 2006; accepted 16 February 2007 Available online 12 April 2007

Abstract

Detritivorous fish generally refers to fish that primarily ingest unidentified organic detritus. We analyzed stomach contents in combination with stable isotopes to trace and compare the food sources of the large-scale mulletLiza macrolepis and other detritivorous fish species in sub-tropical mangrove creeks and a sub-tropical lagoon in Taiwan. The volume of organic detritus always contributed >50% of the stomach content ofL. macrolepis in the two habitats. However, consumed items were distinct between the two habitats and corresponded to the types in which they reside. The consumed items in the lagoon were more diverse than those observed in the mangroves. In the mangroves, the diet composition ofL. macrolepis was primarily determined by season, not by body size. In the lagoon, there were no clear seasonal or size-dependent grouping pat-terns for the diet composition. There were significant seasonal and spatial variations in d13C and d15N values of potential food sources andL. macrolepis. However, neither d13C nor d15N values ofL. macrolepis were correlated with fish body size. Joint analyses of stomach contents and stable isotopes indicated that benthic microalgae on sediments were the most important assimilated food in both seasons for the dominant de-tritivorous fish in the mangroves, whereas a greater reliance on microalgal and macroalgal periphyton on oyster-culture pens was observed in the lagoon. Mangrove and marsh plants and phytoplankton, which are mostly locally produced within each habitat, were of minor importance in the assimilated food.

Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: d13C; d15N; stomach contents; tropical lagoon; mangroves; periphyton

1. Introduction

Many estuarine fish are detritivores, which generally refers to fish that primarily ingest unidentified organic detritus ( Dar-nell, 1961; Ya´~nez-Arancibia, 1976). In estuaries, detritivory:-herbivory ratios are often high when compared to those of other waters (Lin et al., 2001), suggesting that in such environ-ments, organisms are more dependent on organic detritus than directly on plants. The high efficiency of fishery yields (Nixon

et al., 1986) has led researchers to assume that a large propor-tion of estuarine fish are detritivores, which primarily consume the abundant organic detritus and directly convert into fish production (Mann, 1988). However, estuaries are often charac-terized by diverse plant communities and complex origins for the abundant organic detritus they contain. We still do not know food sources of detritivorous fish in estuaries.

Natural stable isotope analyses have been used to trace food sources of carnivorous fish in estuaries. Using d13C,Hughes and Sherr (1983) found that fish in salt marshes were most closely linked to the marsh plantSpartina and benthic algae.

Rodelli et al. (1984)found that gobies in mangrove creeks ob-tained carbon from mangrove trees. Sullivan and Moncreiff

* Corresponding author. Department of Life Sciences, National Chung Hsing University, Taichung 402, Taiwan, ROC.

E-mail address:hjlin@dragon.nchu.edu.tw(H.-J. Lin).

0272-7714/$ - see front matterÓ 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.ecss.2007.02.013

(2)

(1990)measured d13C and d15N and found that the food sour-ces for marsh fish were benthic and planktonic algae. Using d13C and d34S, Stribling and Cornwell (1997) found that C3 plants were of greater importance in the assimilated diets of carnivorous fish in oligohaline wetlands. Combining d13C, d15N, and d34S measurements,Kwak and Zedler (1997)found thatSpartina was the major organic matter source for estuarine fish, whereas macroalgae and microalgae supported fish in a la-goon lackingSpartina.Deegan and Garritt (1997) found that terrestrial organic matter was of minor importance in estuarine food webs.Wainright et al. (2000)found that macrophytes and benthic microalgae were significant components of the diet of Fundulus heteroclitus in salt marshes. It appears that food sources of carnivorous fish in estuaries are very diverse, which might have been attributable to the prey items consisting of a variety of feeding groups and different habitat types. There has been no study to trace the food sources of detritivorous fish in estuaries.

Temporal and spatial variabilities of stable isotope ratios within all organic matter pools are often high (Stephenson et al., 1984; Cloern et al., 2002). These variabilities may ob-scure any source-specific isotopic signatures, confounding the application of stable isotopes for tracing trophic linkages from primary producers to consumers. Stomach content anal-yses can provide a taxonomic resolution of food that may be difficult to achieve by stable isotope analyses. Peterson (1999)suggested that stable isotope analyses are most effec-tively used in combination with other techniques such as stom-ach content analyses. The large-scale mullet Liza macrolepis (family Mugilidae) are the most abundant detritivorous fish in estuaries and mangroves along the western coast of Taiwan (Lin and Shao, 1999; Kuo et al., 2001). In this study, we exam-ined stomach contents in combination with stable carbon and nitrogen isotopes to trace and compare the food sources ofL. macrolepis and other dominant detritivorous fish species in two different habitats: subtropical mangrove creeks draining mangroves and marshes, and a tropical lagoon with fewer vas-cular plants but abundant phytoplankton and periphyton on oyster-culture pens (Fig. 1). We also examined if differences existed between the food sources of L. macrolepis in the dry and wet seasons.

2. Materials and methods 2.1. Study sites

One of the study sites was situated in mangrove creeks of the Guandu wetlands (2570N, 121270E), which is located in an estuary at the confluence of the Danshuei and Keelung Rivers in subtropical northern Taiwan (Fig. 1). These rivers flow through Taipei City of six million people and receive nu-trient inputs from treated and untreated domestic sewage and agricultural effluents. Climatic data derived from a local weather station of the Central Weather Bureau of Taiwan from 1971 to 2000 show that the mean air temperature ranged from 15.1C in January to 28.8C in July. The mean annual and monthly precipitation values were 2120 and 177 mm,

respectively. The Guandu mangroves are composed of the mangroveKandelia candel (ca. 0.55 km2) and marshes con-taining the plants Brachiaria mutica, Clerodendrum inerme, Cyperus malaccensis, Phragmites communis, and Typha an-gustifolia (ca. 0.57 km2). They are subjected to a semidiurnal tidal regime with a tidal amplitude of about 1e2 m. The creeks are about 5 m wide with water about 0.5 m deep covering a thick layer of sandy-silt sediment at low tide. Salinities of the overlying waters ranged from 7 to 12 psu at low tide and might reach 25 psu at high tide. Water column chlorophyll

a remained at 5 mg m3, but could reach 40 mg m3 in

summer (authors’ unpubl. data). Concentrations of dissolved inorganic nitrogen (DIN) in the water column were high (34e195 mM).

The other study site was situated in Tapong Bay (22270N, 120260E), which is a tropical lagoon in southern Taiwan, and has only one tidal inlet connecting it to the sea. Tapong Bay has a 4.44-km2 surface area with a mean depth of 2.2 m at low tide (Fig. 1). It is primarily subjected to semidiurnal tides with a tidal range of 1.0 m. Since no large river flows into the lagoon, the salinity remains high (25e34 psu). Climatic data from 1971 to 2000 of a nearby weather station show that the

mean air temperature ranged from 18.8C in January to

28.9C in July. Despite a mean annual precipitation of 1785 mm, there are distinct dry and wet seasons. During the wet season of MayeSeptember, mean monthly precipitation frequently exceeds 200 mm, and in OctobereApril, mean monthly precipitation normally does not exceed 50 mm. Tapong Bay is lined by the mangroveAvicennia marina and marshes composed ofClerodendrum inerme, Derris laxiflora, andExcoecaria agallocha. The lagoon is densely covered by thousands of oyster-culture pens. Since the lagoon water is often turbid (light extinction > 1.0 m1) and the bottom water

in the inner lagoon possibly becomes hypoxic (Hung and

Hung, 2003), no rooted macrophytes were observed in the la-goon. Benthic microalgae and macroalgal periphyton on oys-ter-culture pens, together with phytoplankton, dominate this lagoon. Water column chlorophylla averages 8.4 mg m3 in summer and 5.8 mg m3in winter (Su et al., 2004). Although there was a wide range of water column DIN concentrations (0.46e23.2 mM), the loads of DIN from surrounding aquacul-ture ponds were high (1.2 mol m2yr1,Lin and Hung, 2004).

2.2. Field collections

Potential food sources, including sestonic particulate or-ganic matter (SPOM), benthic microalgae (mainly diatoms, authors’ pers. obs.), macroalgae, cyanobacteria, marsh plants, and mangrove leaves were collected at both sites. Different body sizes ofLiza macrolepis were collected from both sites for analyses of stomach contents and stable isotopes to exam-ine whether ontogenetic diet changes existed. Other dominant detritivorous fish species, including an Oreochromis hybrid (family Cichlidae),Pomacentrus taeniometopon (Pomacentri-dae),Scarus ghobban (Scaridae), Scatophagus argus

(Scato-phagidae), Siganus guttatus (Siganidae), and Valamugil

(3)

analyses for comparisons. To characterize diet changes be-tween the dry and wet seasons, samples were collected in Ja-nuaryeMarch 2001 (the dry season) and JulyeAugust 2001 (the wet season).

For each sampling of SPOM, 20 L of water was taken 30 cm below the water surface at both flood and ebb tides. Upon returning to the laboratory, water samples were filtered onto precombusted glass fiber filters (GF/F) through a mesh opening diameter of 200 mm, and carbonate was removed by acidification with 1 N HCl. In the mangrove creeks, benthic microalgae appearing as a brown-colored layer on the surface of the sediments were sampled by scraping off 2 mm of the surface layer. Within 1 h of collection, the sediments were brought back to the laboratory where the microalgae were con-centrated and separated from bulk sediments by a procedure modified from that described by Riera and Richard (1996). In the lagoon, benthic microalgae on oyster shells were

washed off in distilled water using a toothbrush and were re-tained on precombusted glass fiber filters (GF/F) by sieving through a mesh with an opening diameter of 62 mm. Senescent leaves of mangrove and marsh plants in the mangrove creeks and macroalgal periphyton on oyster-culture pens in the la-goon were, respectively, collected and vigorously washed with distilled water to remove epiphytes and extraneous sedi-ments. These samples of potential food sources were then freeze-dried and ground for stable isotope analyses.

Fish were collected using fyke nets. This fishing gear is a passive sampler designed to collect nekton in a depth range of 0.5e2.0 m and is composed of two fence nets (15 m long and 1.5 m high, with a mesh opening diameter of 15 mm) and a hoop-net (with a mesh opening diameter of 10 mm). Fyke nets were set out and uplifted early in the morning (04:00) when most fish were feeding. For each captured fish, total length (TL) and fresh weight were recorded, and white

Taiwan Strait

Tapong Bay

The Guandu wetlands

121° 122° 22° 23° 24° 25° Pacific Ocean 120° barrier Lipan Creek Mangrove Creek 1 km I l t N Inlet Aquaculture pond Mangroves Taiwan Strait Gueytzykeng Creek Chungkang Creek Shuimokeng Creek ditch gate

ababdoned rice paddies

pond pond pond dyke tidal creek N 100 m

TAIWAN Tapong Bay

The Guandu mangroves The Guandu marshes

(4)

muscle tissue and digestive tract were removed and frozen for later analyses of stomach content and stable isotopes, respec-tively. Because of budget constraints and integrative nature of stable isotopes, fish samples subjected to examination of stom-ach content were randomly assigned for stable isotope analyses.

2.3. Stomach content analyses

Stomach contents were sorted taxonomically and counted using a light microscope at 400 after sedimentation on a slide. Two transects (about 100 fields) were scanned on a slide. The frequency of occurrence method was employed by recording the number of stomachs containing one or more individuals of each consumed item, and the total was ex-pressed as a percentage of the total number of stomachs exam-ined (FO%). The volume of a consumed item was then determined by squashing the stomach contents on a slide to a uniform depth and the area of the squash was measured. The volume of a consumed item taken by a fish population was given as a percentage of the total volume of the stomach content (V%). Consideration of both volumetric percentage and frequency of occurrence percentage for a consumed item can provide an indication of the homogeneity of feeding within a fish population (Hyslop, 1980). Fish samples with empty stomachs were excluded from this analysis.

2.4. Stable isotope analyses

We analyzed d13C and d15N of white muscle tissues of the fish. Each fish sample was treated individually. d13C values were used to determine carbon sources. d15N values were an-alyzed as a secondary tracer to indicate the trophic position. Tissues were freeze-dried and then ground. Because the 13C content of lipids may be depleted and this can affect ecological interpretations (Kling et al., 1992), samples were washed in a 2:1:0.8 methanol:chloroform:water solution for 2 h (Bligh and Dyer, 1959), and then treated with 1 N HCl to remove lipids and carbonates (Boutton, 1991). The d13C and d15N values of samples were determined with a continuous-flow iso-tope ratio mass spectrometer (Finnigan deltaS ) coupled with an elemental analyzer (Carlo Erba NA 1500 NCS). Because lipid extraction and acidification have been found to alter tis-sue d15N (Pinnegar and Polunin, 1999; Sweeting et al., 2006), chemically treated and untreated samples were run for d13C and d15N, respectively. The precision of the measurements was 0.1& for both the stable carbon and nitrogen isotope analyses. Stable isotope data were reported as the relative dif-ference between ratios of a sample and standards in standard notation as: dX (&)¼ [(Rsample/Rstandard) 1]  103, where R¼13C/12C or15N/14N. d13C or d15N is the per-mil (&) devi-ation of that sample from the recognized isotope standard, i.e., PeeDee Belemnite (PDB) limestone for d13C and atmospheric N2 for d15N (Gonfiantini et al., 1995). DeNiro and Epstein

(1978) found that an animal is on average enriched in d13C by about 1& relative to its diet.Minagawa and Wada (1984)

andPost (2002)showed that the mean enrichment in d15N at

a single feeding process is 3.4&. The dual-isotope, multiple-source mixing model (Ben-David et al., 1997), was then used as an index of relative contribution of each source to the assimilated diet of the dominant detritivorous fish species in each habitat.

2.5. Statistical analyses

In order to reveal seasonal and ontogenetic patterns in the diet ofLiza macrolepis in each habitat, changes in the compo-sition of stomach contents were studied at each site using mul-tivariate analyses in the PRIMER (vers. 5.2) computer package (Clarke and Gorley, 2001). BrayeCurtis coefficient was used to produce a similarity matrix of volumetric compo-sition of stomach contents between any two samples. Before calculating the similarity coefficients, the volumetric percent-age of each food item was square root-transformed using the power transformations (Clarke and Warwick, 1994) to down-weight the effects of dominant consumed items. The similarity matrix was classified by hierarchical agglomerative clustering using the unweighted pair group mean arithmetic (UPGMA) linking method. A Chi-square test was then used to test differ-ences in volumetric composition of consumed items ofL. mac-rolepis between the dry and wet seasons.

For each site, Spearman’s rank correlations (rs) between d13C and d15N and total length ofLiza macrolepis were com-puted using data from individual fish to examine whether on-togenetic diet changes existed. A two-way fixed ANOVA model was then used to evaluate whether d13C and d15N values of potential food sources and of L. macrolepis differed be-tween the two seasons (the dry and wet seasons) or the two sites (the mangrove creeks and the tropical lagoon). Before the analyses, the data were tested by the power transforma-tions for normality and homogeneity of variance assumptransforma-tions (Clarke and Warwick, 1994). These univariate statistical cal-culations were produced using the SAS system (vers. 8.2). 3. Results

3.1. Stomach content analyses

In the Guandu mangroves and Tapong Bay, the sampled

body sizes of Liza macrolepis ranged 39e242 mm, which

are generally beyond its maturity size of 40 mm total length (Froese and Pauly, 2006). In the Guandu mangroves, classifi-cation of the volumetric composition (V%) of consumed items from each individual fish ofL. macrolepis separated the stom-ach contents into two major categories (D and W) at a similar-ity level of 65% (Fig. 2a). Category D contained stomach contents mainly from the dry season, while category W was composed of samples from the wet season. The grouping pat-terns showed that the diet composition ofL. macrolepis in the Guandu mangroves was primarily determined by season, not by body size.

In the mangroves, unidentified organic detritus, epipelic di-atoms (mainly Navicula and Nitzschia, Round et al., 1990), and vascular plants were the most frequently consumed items

(5)

(FO%) in the stomach contents ofL. macrolepis (Table 1) in both seasons. Volumetric compositions (V%) of the stomach contents significantly differed between the two seasons (Chi-square test, p < 0.001). The contributions of organic detritus and vascular plants were co-dominant in the dry season, whereas in the wet season, the percentages of organic detritus and diatoms increased, while that of vascular plants decreased. In Tapong Bay, there were no clear seasonal or size-dependent grouping patterns for volumetric compositions of the stomach contents of Liza macrolepis (Fig. 2b). Unidenti-fied organic detritus, macroalgae, and epipelic diatoms

(mainly Navicula) were the dominant consumed items and

together contributed >85% to the total volume (Table 1). Ben-thic invertebrates (snails and amphipods) were also observed in the stomach contents, but the V% values were small.

Consumed items in the stomach contents of Liza macrole-pis in the tropical lagoon were more diverse than those ob-served in the mangrove creeks (Table 1). In addition to the consumed items recorded in the mangroves, dinoflagellates, benthic invertebrates, and fungi were recorded from fish in the lagoon. Volumetric compositions of consumed items dif-fered between the two habitats (Chi-square test, p < 0.001). In the lagoon, the contribution of macroalgae and diatoms was greater, but that of vascular plants was lower.

46 W 50 D 46 W 47 W 55 W 45 W 45 W 48 W 47 D 47 W 53 W 46 D 60 D 39 D 55 D 55 W 48 D 49 D 51 D 43 D 56 D 40 60 80 100 40 60 80 100 180 D 66 W 75 W 95 D 93 D 100 W 160 D 87 D 58 W 92 D 100 D 62 W 155 D 40 W 143 D 162 D 147 W 155 D 155 D 72 W 55 W 86 W 181 D 89 D 82 W 166 D 58 W 180 D 83 D 165 W 200 D 97 D 98 D 95 D 61 W Bray-Curtis similarity (%) D W b a

Fig. 2. Classification of the diet composition of individual fish ofLiza macrolepis with different total lengths (in mm) collected from (a) the Guandu mangroves and (b) Tapong Bay in the dry (D) and wet (W) seasons.

(6)

3.2. Stable isotope analyses

d13C and d15N values of potential food sources in the Guandu mangroves and Tapong Bay showed clear spatial and seasonal variations (Tables 2 and 3). Generally, d13C and d15N values of the same functional groups were more en-riched in the tropical lagoon than in the mangroves (Table 4). At both sites, d13C values of SPOM were more enriched in the dry season than in the wet season. However, there was an in-teractive effect of site and season on d13C values of benthic microalgae, which were more enriched in the wet season in Tapong Bay, but were more enriched in the dry season in the mangroves. d15N values of benthic microalgae showed en-richment in the wet season at both sites.

Different species of detritivorous fish collected from each habitat in the same season generally had similar d13C values (Tables 5 and 6). Size-dependent variations in d13C (Spear-man’s rank correlation, rs¼ 0.018 and p ¼ 0.96 in the

Guandu mangroves; rs¼ 0.028, p ¼ 0.88 in Tapong Bay)

and d15N (Spearman’s rank correlation, rs¼ 0.379,

p¼ 0.25 in the Guandu mangroves; rs¼ 0.028, p ¼ 0.88 in Tapong Bay) values of Liza macrolepis were not significant within the size range sampled in each habitat. Like spatial var-iations of potential food sources, d13C and d15N values of L. macrolepis were also more enriched in Tapong Bay than in the Guandu mangroves (Table 4). In the mangroves, d13C and d15N values were more enriched inL. macrolepis collected in the dry season than in the wet season, although the interac-tive effect of site and season on d13C was marginally

signifi-cant (two-way ANOVA, p¼ 0.10).

According to the trophic fractionation of C and N (DeNiro and Epstein, 1978; Minagawa and Wada, 1984; Post, 2002), in the Guandu mangroves, data points of the d13C vs. d15N plots for the dry (Fig. 3a) and wet seasons (Fig. 3b) appeared that benthic microalgae on sediments and a mixture of C4marsh plants and SPOM were the primary food sources for the dom-inant detritivorous fish. However, d13C values of benthic mi-croalgae and detritivorous fish shifted in the same direction seasonally, from higher values in the dry season to lower values in the wet season, whereas a slight shift occurred on the values of C4 marsh plants. In addition, stomach content analyses showed that the dominant genera of diatoms in the di-ets ofLiza macrolepis were epipelic Navicula and Nitzschia. This suggested that benthic microalgae were the main food sources for the dominant detritivorous fish. The mixing model further indicated that the relative contribution of benthic mi-croalgae to the assimilated food averaged 45%.

In Tapong Bay, data points of the d13C vs. d15N plots for the dry (Fig. 4a) and wet seasons (Fig. 4b) indicated a great reli-ance of the dominant detritivorous fish on benthic microalgae and macroalgal periphyton on oyster-culture pens. The mixing model showed that the summed percentage of benthic micro-algae and macroalgal periphyton in the assimilated food was consistently >55%.

4. Discussion

C3marsh plants and SPOM collected from Tapong Bay and the Guandu mangroves showed a clear spatial variation in d13C

Table 1

Frequency of occurrence (FO%) and percentage by volume (V%) of consumed items in the diets ofLiza macrolepis in the Guandu mangroves and Tapong Bay collected in the wet and dry seasons of 2001. Sample size (n) refers to the total number of stomachs examined; also given under the size class are the means SE, the range of total lengths, and the percentages of stomachs that were empty

Site/season Size class Item FO% V% The Guandu mangroves

Dry season n¼ 36 Cyanobacteria 18.2 0.2 49.5 1.8 mm Diatoms 90.9 4.4 39e60 mm Macroalgae 9.1 0.1 53% empty Vascular plants 100 43.2 Organic detritus 100 52.1 Wet season n¼ 32 Cyanobacteria 20.0 0.3 48.7 1.3 mm Diatoms 100 8.1 43e49 mm Macroalgae 60.0 0.3 41% empty Vascular plants 80.0 11.2 Organic detritus 100 80.1 Tapong Bay n¼ 72 Cyanobacteria 51.4 2.9 111.2 7.7 mm Diatoms 100 10.8 40e200 mm Dinoflagellates 45.7 0.8 29% empty Macroalgae 88.6 19.1 Fungi 14.1 0.2 Vascular plants 54.3 5.3 Benthic invertebrates 34.1 4.5 Organic detritus 100 56.1 Table 2

Summary of d13C (&) and d15N (&) of potential food sources collected in the

Guandu mangroves. Means SE and the individual number measured (n) are shown below. SPOM, sestonic particulate organic matter

Season Food source (n) d13C (&) d15N (&)

Dry C4marsh plants

Cyperus malaccensis (1) 12.7 6.3 Brachiaria mutica (1) 12.9 6.3 C3marsh plants Phragmites communis (2) 27.4  0.6 5.6 1.5 Clerodendrum inerme (1) 26.9 6.7 Typha angustifolia (1) 28.1 4.5 MangroveKandelia candel (1) 29.6 7.2 Benthic microalgae (3) 22.5  0.1 2.3 0.2 SPOM at flood tide (2) 23.6  0.1 2.7 0.2 SPOM at ebb tide (2) 24.7  0.1 0.5  0.1 Wet C4marsh plants

Brachiaria mutica (1) 13.6 6.2 C3marsh plants

Phragmites communis (1) 26.9 1.4 Clerodendrum inerme (1) 27.9 5.8 Typha angustifolia (1) 29.1 4.4 MangroveKandelia candel (1) 28.2 6.8 Benthic microalgae (2) 24.1  0.1 3.4 0.3 SPOM at flood tide (2) 26.1  0.1 0.02  0.01 SPOM at ebb tide (2) 27.1  0.1 2.4  0.1

(7)

values, which were more enriched in the tropical lagoon than in the subtropical mangroves. A possible cause of this spatial pattern is the different sources of dissolved inorganic carbon between the two sites and isotopic fractionation by plants. The d13C values of terrestrial and freshwater C3plants or oli-gohaline phytoplankton have been shown to be more depleted than the values of marine benthic diatoms or phytoplankton along a salinity gradient in estuaries (e.g. Fry and Sherr, 1984; Hsieh et al., 2000). While Tapong Bay receives little riv-erine input, the Guandu mangroves are greatly influenced by the discharge of the Danshuei and Keelung Rivers and the d13C signatures are thus more depleted. This is also evidenced by the clear seasonal patterns for the diet composition ofLiza macrolepis from the mangroves, but not from the lagoon. Other possibility is water use efficiency of plants. Farquhar et al. (1989)found that there is a negative linear relationship between d13C values of C3leaves and the ratios of intercellular to ambient CO2concentrations, implying that d13C values can be used to indicate the long-term integrated water use effi-ciency of plants. The mean annual and monthly precipitation values were lower in Tapong Bay and were concentrated only within the 2e3 months of summer. In addition, since no large river flows into the lagoon, the salinity in the lagoon is remarkably higher than in the mangroves. It is likely that the

enriched d13C values of C3marsh plants taken from the lagoon are due to less water being available and to the maintenance of a higher water use efficiency. Salinities of intertidal sediments taken from the mangroves were often higher than those of overlying water (Lin et al., 2003). This may also explain the higher d13C values of C4marsh plants and benthic microalgae on intertidal sediments taken from the Guandu mangroves in response to less water availability in the dry season.

In Tapong Bay, however, d13C values of benthic microalgae and macroalgal periphyton on oyster-culture pens submerged in the water column were more enriched in the wet season than in the dry season.Finlay et al. (2002)observed that algal d13C in riffles increased during the summer when pH was highest and CO2 was lowest in the water column. They also found a strong negative relationship between algal d13C and water velocity in rivers. In Tapong Bay, productivity and bio-mass of phytoplankton and pH are greater in the wet season (Hung and Hung, 2003; Su et al., 2004). Therefore, the higher algal d13C measured in the lagoon in the wet season may have been due to changes in pH or water turbulence resulting in changes in CO2in the water column.

d15N values of terrestrial plants (C3and C4marsh plants) did not appear to be affected by site or season. Nevertheless, d15N values of algae and SPOM showed a significant spatial variation, which were more enriched in Tapong Bay than in the Guandu mangroves. This suggests that the heterogeneity of d15N values was relevant to N sources in the water column. d15N values have been useful in detecting anthropogenic N loading. d15N values are often high for nutrient inputs of sew-age and aquaculture ponds but low for agricultural inputs (McClelland et al., 1997; Fry, 1999; Jones et al., 2001; Cos-tanzo et al., 2003). In Tapong Bay, the loads of sewage DIN (1.2 mol m2yr1) from surrounding aquaculture ponds were high when compared to those in other estuaries and coastal lagoons (Lin and Hung, 2004). For the dominant pe-riphyton species Ulva lactuca, Enteromorpha intestinalis, andLyngbya majuscula in the lagoon, mean tissue N reached 4.1e6.5% dry wt., which is comparable to values of similar taxa found in other studies under enriched conditions (Kamer et al., 2001). Accordingly, the elevated d15N and %N of algae collected in Tapong Bay are indicative of sewage inputs from surrounding aquaculture ponds. The lower d15N values in the Guandu mangroves suggest that the Danshuei and Keelung Rivers receive both nutrient inputs from domestic sewage and agricultural effluents.

Mullet are often the dominant fish species in estuaries and coastal shallow waters worldwide. Food items reported in the stomach contents of mullet are primarily diatoms, macroalgae, and unidentified organic detritus (Odum, 1970; Ya´~ nez-Aranci-bia, 1976; Chan and Chua, 1979). However, there is substan-tial heterogeneity in the relative importance of each item in

the stomach contents. As a result, Odum (1970) and Chan

and Chua (1979)considered mullet to be herbivores, whereas

Ya´~nez-Arancibia (1976) reported that mullet are detritivores. In this study, consumed items in the stomach contents of Liza macrolepis were distinct between the two habitats and corresponded to the types in which they reside. In the Guandu

Table 3

Summary of d13C (&) and d15N (&) of potential food sources collected in

Tapong Bay. Means SE and the individual number measured (n) are shown below. SPOM, sestonic particulate organic matter

Season Food source (n) d13C (&) d15N (&)

Dry C3marsh plants

Clerodendrum inerme (1) 23.7 9.9 Excoecaria agallocha (1) 25.5 10.1 Derris laxiflora (1) 27.6 1.4 MangroveAvicennia marina (1) 26.3 5.6 Macroalgal periphyton Ceramium cimbricum (3) 10.6  0.1 10.9 0.1 Chaetomorpha crassa (5) 11.4  0.7 10.4 0.1 Enteromorpha intestinalis (30) 11.7  0.6 11.1 0.1 Ulva fasciata (9) 11.3  0.7 11.4 0.8 Ulva lactuca (22) 11.2  0.3 11.0 0.2 Lyngbya majuscula (24) 12.7  0.6 10.5 0.2 Benthic microalgae (3) 22.9  0.1 11.6 0.1 SPOM at ebb tide (1) 19.0 7.0 Wet C3marsh plants

Clerodendrum inerme (1) 27.7 4.4 Excoecaria agallocha (1) 26.6 12.0 Derris laxiflora (1) 26.2 5.5 MangroveAvicennia marina (1) 28.1 10.5 Macroalgal periphyton Enteromorpha intestinalis (12) 11.5  0.8 10.2 0.4 Ulva fasciata (3) 10.4  0.3 11.5 0.8 Ulva lactuca (2) 10.8  0.2 10.6 0.1 Lyngbya majuscula (5) 14.9  1.9 10.7 0.3 Benthic microalgae (2) 17.8  0.2 12.7 0.2 SPOM at flood tide (1) 23.4 5.4 SPOM at ebb tide (1) 23.5 7.1

(8)

mangroves with abundant vascular plants, the contribution of vascular plants to the volume of the stomach content was high and reached >40% in the dry season. In the more produc-tive Tapong Bay (Lin et al., 2006), consumed items in the stomach contents were more diverse than those observed in the mangroves. In the lagoon densely covered by thousands of oyster-culture pens, the contribution of benthic microalgae and macroalgal periphyton to the volume of the stomach con-tent reached 30%. Although the contribution to the volume of

the stomach content was <5%, benthic invertebrates were ob-served in 34% of the stomach contents taken in Tapong Bay.

Deegan and Garritt (1997)indicated that estuarine consumers tend to utilize organic sources produced in the same region in which they reside. Hence, the spatial heterogeneity in stomach contents of mullet may be attributable to the fact that the mul-let examined in previous studies were taken from different habitat types. Despite this, the volume of organic detritus al-ways contributed >50% of the stomach content ofL. macro-lepis in the two distinct habitats (Table 1). Our results suggest that the large-scale mullet are detritivores.

Despite the diverse diets and the distinct consumed items between Tapong Bay and the Guandu mangroves, d15N values of Liza macrolepis and other dominant detritivorous fish in these two habitats indicated that these fish are feeding approx-imately one trophic level above primary producers. Joint anal-yses of stomach contents and d13C values demonstrated that the assimilated food of these fish taken from the two different habitats was benthic algae and corresponded to the dominant taxa in which they reside (Figs. 2 and 3). In the Guandu man-groves, L. macrolepis assimilated mainly benthic microalgae on sediments, whereas in Tapong Bay, microalgal and macro-algal periphyton on oyster-culture pens were the most impor-tant food sources. In mangrove creeks such as the Guandu mangroves, the availability of organic carbon from mangroves is often high (Rezende et al., 1990). The stomach contents of L. macrolepis in the mangroves also showed a high proportion

Table 4

Two-way ANOVA for the d13C (&) and d15N (&) values of potential food sources andLiza macrolepis collected in the dry and wet seasons from the Guandu

mangroves (G) and Tapong Bay (T ). *p < 0.05. SPOM, sestonic particulate organic matter

Source/fish d13C d15N df MS F p df MS F p C3marsh plants Site 1 6.78 4.97 0.05* 1 17.9 1.77 0.21 Season 1 2.36 1.74 0.21 1 2.30 0.18 0.68 Site season 1 0.40 0.28 0.61 1 2.89 0.26 0.62 Error 9 1.42 9 11.1

Result G < T, Dry¼ Wet G¼ T, Dry ¼ Wet Benthic microalgae

Site 1 21.1 649 <0.001* 1 211 2925 <0.001*

Season 1 7.26 223 <0.001* 1 2.65 36.6 0.001*

Site season 1 27.2 838 <0.001* 1 0.001 0.01 0.93

Error 6 0.03 6 0.07

Result G < T in the wet season, Dry > Wet in G, Dry < Wet in T G < T, Dry < Wet SPOM

Site 1 28.2 87.7 0.001* 1 52.1 21.6 0.01*

Season 1 21.7 67.4 0.001* 1 0.08 0.09 0.78

Site season 1 1.28 3.99 0.12 1 0.43 0.19 0.68

Error 4 0.32 4 2.31

Result G < T, Dry > Wet G < T, Dry¼ Wet Liza macrolepis

Site 1 189 24.1 <0.001* 1 1186 214 <0.001*

Season 1 8.79 1.12 0.30 1 17.3 3.14 0.09

Site season 1 22.6 2.88 0.10 1 22.5 4.07 0.05*

Error 38 7.85 38 5.52

Result G < T, Dry¼ Wet G < T in both seasons, Dry > Wet in G

Table 5

Summary of d13C (&) and d15N (&) of the dominant detritivorous fish

col-lected in the Guandu mangroves. Means SE and the individual number measured (n) are shown below

Season Fish species d13C (&) d15N (&) Dry Liza macrolepis

38.0 2.0 mm (2) 18.7  0.2 8.5 0.9 63.0 6.5 mm (3) 19.6  0.3 2.3 0.6 Oreochromis hybrid

95 mm (1) 21.5 4.7 Wet Liza macrolepis

37.5 1.3 mm (2) 22.2  0.1 1.7 0.3 46.5 0.5 mm (2) 21.7  0.2 1.8 0.3 55.0 0.0 mm (2) 22.1  0.2 1.2 0.7 Oreochromis hybrid

(9)

of vascular plants. However, our results of stable isotope anal-yses demonstrated that vascular plants were little assimilated and utilized by the large-scale mullet. The reason for the rel-atively low proportion of algae observed in the stomach con-tent might be attributable to the fast digestion rates. Softer and more easily digested items, such as algae, might persist in the stomach for much less time than refractory macrophyte detritus. The organic detritus most often occurring in the stom-ach contents was likely unidentified algal detritus and/or re-fractory food which is difficult to digest.

In Tapong Bay, high nutrient loadings and long water resi-dence times (up to 24 d) in the lagoon stimulate the growth and accumulation of phytoplankton and periphyton on oys-ter-culture pens. The abundance and productivity of phyto-plankton and periphyton in the lagoon were higher than those recorded in most estuaries and coastal lagoons (Lin and Hung, 2004). The relative contribution of phytoplankton to system productivity and biomass was much more important than that of periphyton in the lagoon. Periphyton biomass might exceed that of phytoplankton only in winter and spring when periphyton bloom (Lin et al., 2005). The high chloro-phyll a contents (mean: 7.1 mg m3, Su et al., 2004) and lowC:N ratios (mean: 7.6,Hung and Hung, 2003) were indic-ative of phytoplankton-rich SPOM (c.f. Cloern et al., 2002). Despite the dominance of phytoplankton in Tapong Bay, joint analyses of stomach contents and stable isotopes revealed that

SPOM were of minor importance in the food sources of L. macrolepis in the lagoon, regardless of season. Sobczak et al. (2002) found that bioavailable POM derived primarily from internal phytoplankton production was the dominant food supply to the food web in a temperate estuary. However, our results suggest that in Tapong Bay, there is not a tight cou-pling between pelagic primary production and the food sour-ces of dominant detritivorous fish. In our opinion, a shift in the food chain from phytoplankton-grazing to periphyton-grazing organisms has likely occurred in the tropical lagoon. While man-made aquaculture structures were being added to Tapong Bay, periphyton blooms on the introduced substratum might have altered the trophic pathways through which organic matter is transferred within the food web.

5. Conclusions

Despite the diverse diets and the distinct consumed items of the large-scale mullet Liza macrolepis between Tapong Bay and the Guandu mangroves, the volume of organic detritus al-ways contributed >50% of the stomach content, suggesting that they are detritivores. However, joint analyses of stomach contents and stable isotopes indicated that benthic microalgae on sediments were the most important assimilated food for the

Table 6

Summary of d13C (&) and d15N (&) of the dominant detritivorous fish

col-lected in Tapong Bay. Means SE and the individual number measured (n) are shown below

Season Fish species d13C (&) d15N (&)

Dry Liza macrolepis

92.0 2.3 mm (5) 11.9  0.3 17.5 0.1 156.3 3.5 mm (3) 21.2  0.6 15.3 0.1 209.5 2.5 mm (2) 15.2  0.1 14.2 0.1 242 mm (1) 23.1 6.9 Pomacentrus taeniometopon 91.7 1.8 mm (3) 15.2  0.1 13.3 0.2 Scarus ghobban 179.0 26.0 mm (2) 16.0  0.1 12.5 0.2 Scatophagus argus 110.0 10.0 mm (2) 16.9  1.1 12.6 0.5 Siganus guttatus 119.3 2.6 mm (3) 13.0  1.6 14.6 0.9 Valamugil cunnesius 184.3 6.4 mm (3) 13.8  1.2 14.1 0.6 Wet Liza macrolepis

60.0 2.9 mm (3) 14.8  0.5 17.1 0.1 66.0 0.0 mm (2) 19.1  0.2 14.1 0.4 77.5 0.5 mm (2) 15.2  0.7 13.7 0.1 92.5 2.5 mm (2) 14.5  0.7 17.4 0.1 98.0 6.0 mm (2) 15.2  0.4 12.9 0.7 107.5 0.5 mm (2) 14.8  1.0 13.5 0.5 152.5 2.5 mm (2) 12.2  0.8 17.8 0.1 173.0 0.0 mm (2) 16.8  0.5 16.4 0.2 211.3 6.8 mm (3) 15.8  0.5 16.0 0.1 O. hybrid -5 0 5 10 15 20 L. macrolepis δ 15 N (‰) δ 15 N (‰) δ13C (‰) SPOM C4Marsh plants C3Marsh plants Mangrove leaf Benthic microalgae O. hybrid -5 0 5 10 15 20 -30 -25 -20 -15 -10 L. macrolepis SPOM C4Marsh plants C3Marsh plants Mangrove leaf Benthic microalgae The Guandu mangroves

Dry

Wet a

b

Fig. 3. d15N vs. d13C of food sources and dominant detritivorous fish collected

from the Guandu mangroves in the dry (a) and wet (b) seasons. Boxes indicate ranges.

(10)

dominant detritivorous fish in the mangroves, whereas a greater reliance on microalgal and macroalgal periphyton on oyster-culture pens was observed in the tropical lagoon. Mangrove and marsh plants and phytoplankton, which are mostly locally produced within each habitat, respectively, were of minor im-portance in the assimilated food.

Acknowledgements

This study was supported by the National Science Council under grant number NSC89-2321-B-005-313. The authors thank Dr. Kwang-Tsao Shao and Dr. Hwey-Lian Hsieh at Re-search Center for Biodiversity, Academia Sinica for their help in collecting fish specimens and sample processing.

References

Ben-David, M., Flynn, R.W., Schell, D.M., 1997. Annual and seasonal changes in diets of martens: evidence from stable isotope analysis. Oeco-logia 111, 280e291.

Bligh, E.G., Dyer, W.J., 1959. A rapid method of total lipid extraction and purification. Canadian Journal of Biochemistry and Physiology 37, 911e917. Boutton, T.W., 1991. Stable carbon isotope ratios of natural material: I. Sam-ple preparation and mass spectrometric analysis. In: Coleman, D.C., Fry, B. (Eds.), Carbon Isotope Techniques. Academic Press, New York, pp. 155e171.

Chan, E.H., Chua, T.E., 1979. The food and feeding habits of greenback grey mullet,Liza subviridis (Valenciennes), from different habitats and at vari-ous stages of growth. Journal of Fish Biology 15, 165e171.

Clarke, K.R., Warwick, R.M., 1994. Changes in Marine Communities: An Ap-proach to Statistical Analysis and Interpretation. Natural Environment Re-search Council, Plymouth, UK, 144 pp.

Clarke, K.R., Gorley, R.N., 2001. PRIMER v5: User Manual/Tutorial. PRIMER-E, Plymouth, UK, 91 pp.

Cloern, J.E., Canuel, E.A., Harris, D., 2002. Stable carbon and nitrogen iso-tope composition of aquatic and terrestrial plants of the San Francisco Bay estuarine system. Limnology and Oceanography 47, 713e729. Costanzo, S.D., O’Donohue, M., Dennison, W.C., 2003. Assessing the

sea-sonal influence of sewage and agricultural nutrient inputs in a subtropical river estuary. Estuaries 26, 857e865.

Darnell, R.M., 1961. Tropical spectrum of an estuarine community, based on studies of lake Pontchartrain, Louisiana. Ecology 42, 553e568. Deegan, L.A., Garritt, R.H., 1997. Evidence for spatial variability in estuarine

food webs. Marine Ecology Progress Series 147, 31e47.

DeNiro, M.J., Epstein, S., 1978. Influence of diet on the distribution of carbon isotopes in animals. Geochimica et Cosmochimica Acta 42, 495e506. Farquhar, G.D., Hubick, K.T., Condon, A.G., Richard, R.A., 1989. Carbon

iso-tope fractionation and plant water-use efficiency. In: Rundel, P.W., Ehleringer, J.R., Nagy, K.A. (Eds.), Stable Isotopes in Ecological Re-search. Springer Verlag, Berlin, pp. 21e41.

Finlay, J.C., Khandwala, S., Power, M.E., 2002. Spatial scales of carbon flow in a river food web. Ecology 83, 1845e1859.

Froese, R., Pauly, D., 2006. FishBase. World Wide Web electronic publication. Avaialble from:<www.fishbase.org>. Version (03/2006).

Fry, B., 1999. Using stable isotopes to monitor watershed influences on aquatic trophodynamics. Canadian Journal of Fisheries and Aquatic Sciences 56, 2167e2171.

Fry, B., Sherr, E.B., 1984. d13C measurements as indicators of carbon flow in marine and freshwater ecosystems. Contributions in Marine Science 27, 13e47.

Gonfiantini, R., Stichler, W., Rozanski, K., 1995. Standards and intercompar-ison materials distributed by the international atomic energy agency for stable isotope measurements. In: References and Intercomparison Mate-rials for Stable Isotopes of Light Elements. IAEA, Vienna, pp. 13e29 (IAEA-TECDOC-825).

Hsieh, H.-L., Kao, W.-Y., Chen, C.-P., Liu, P.-J., 2000. Detrital flows through the feeding pathway of the oyster (Crassostrea gigas) in a tropical shallow lagoon: d13C signals. Marine Biology 136, 677e684.

Hughes, E.H., Sherr, E.B., 1983. Subtidal food webs in a Georgia estuary: d13C analysis. Journal of Experimental Marine Biology and Ecology 67,

227e242.

Hung, J.-J., Hung, P.-Y., 2003. Carbon and nutrient dynamics in a hypertrophic lagoon in southwestern Taiwan. Journal of Marine System 42, 97e114. Hyslop, E.L., 1980. Stomach contents analysis e a review of methods and

their application. Journal of Fish Biology 17, 411e429.

Jones, A.B., O’Donohue, M.J., Udy, J., Dennison, W.C.D., 2001. Assessing ecological impacts of shrimp and sewage effluent: biological indicators with standard water quality analyses. Estuarine, Coastal and Shelf Science 52, 91e109.

Kamer, K., Boyl, K.A., Fong, P., 2001. Macroalgal bloom dynamics in a highly eutrophic southern California estuary. Estuaries 24, 623e635.

Kling, G.W., Fry, B., O’Brien, W.J., 1992. Stable isotopes and planktonic tro-phic structure in Arctic lakes. Ecology 73, 561e566.

Kuo, S.-R., Lin, H.-J., Shao, K.-T., 2001. Seasonal changes in abundance and composition of the fish assemblage in Chiku Lagoon, Southwestern Tai-wan. Bulletin of Marine Science 68, 85e99.

Kwak, T.J., Zedler, J.B., 1997. Food web analysis of southern California coastal wetlands using multiple stable isotopes. Oecologia 110, 262e277. Lin, H.-J., Dai, X.-X., Shao, K.-T., Su, H.-M., Lo, W.-T., Hsieh, H.-L., Fang, L.-S., Hung, J.-J., 2006. Trophic structure and functioning in a eutro-phic and poorly-flushed lagoon in southern Taiwan. Marine Environmental Research 62, 61e82.

Lin, H.-J., Wang, T.-C., Su, H.-M., Hung, J.-J., 2005. Relative importance of phytoplankton and periphyton on oyster-culture pens in a eutrophic tropi-cal lagoon. Aquaculture 243, 279e290.

Lin, H.-J., Hung, J.-J., 2004. Factors affecting macroalgal distribution in a eu-trophic tropical lagoon in Taiwan. Marine Biology 144, 653e664.

Tapong Bay Dry Wet L. macrolepis SPOM C3 Marsh plants Mangrove leaf

other dominat detritivous fishes

Macroalgal periphyton Benthic microalgae L. macrolepis SPOM C3 Marsh plants Mangrove leaf Macroalgal periphyton Benthic microalgae a b -5 0 5 10 15 20 δ 15 N (‰) δ 15 N (‰) δ13C (‰) -5 0 5 10 15 20 -30 -25 -20 -15 -10

Fig. 4. d15N vs. d13C of food sources and dominant detritivorous fish collected from Tapong Bay in the dry (a) and wet (b) seasons. Boxes indicate ranges.

(11)

Lin, H.-J., Shao, K.-T., Chiou, W.-L., Maa, C.-J.W., Hsieh, H.-L., Wu, W.-L., Severinghaus, L.L., Wang, Y.-T., 2003. Biotic communities of freshwater marshes and mangroves in relation to saltwater incursions: implications for wetland regulation. Biodiversity and Conservation 12, 647e665. Lin, H.-J., Hung, J.-J., Shao, K.-T., Kuo, F., 2001. Trophic functioning and

nu-trient flux of a highly productive tropical lagoon. Oecologia 129, 395e406. Lin, H.-J., Shao, K.-T., 1999. Seasonal and diet changes in a subtropical

man-grove fish assemblage. Bulletin of Marine Science 65, 775e794. Mann, K.H., 1988. Production and use of detritus in various freshwater, estuarine,

and coastal marine ecosystems. Limnology and Oceanography 33, 910e930. McClelland, J.W., Valiela, I., Michener, R.H., 1997. Nitrogen-stable isotope signatures in estuarine food webs: a record of increasing urbanization in coastal watersheds. Limnology and Oceanography 42, 930e937. Minagawa, M., Wada, E., 1984. Stepwise enrichment of 15N along food

chains: further evidence and the relation between d15N and animal age. Geochimica et Cosmochimica Acta 48, 1135e1140.

Nixon, S.W., Oviatt, C.A., Frithsen, J., Sullivan, B., 1986. Nutrients and the productivity of estuarine and coastal ecosystems. Journal of Limnological Society of Southern Africa 12, 43e71.

Odum, W.E., 1970. Utilization of the direct grazing and plant detritus food chains by the striped mulletMugil cephalus. In: Steele, J.H. (Ed.), Marine Food Chains. Oliver & Boyd, Edinburgh, pp. 222e240.

Peterson, B., 1999. Stable isotopes as tracers of organic matter input and trans-fer in benthic food webs: a review. Acta oecologica 20, 479e487. Pinnegar, J.K., Polunin, V.C., 1999. Differential fractionation of d13C and

d15N among fish tissues: implications for the study of trophic interactions. Functional Ecology 13, 225e231.

Post, D.M., 2002. Using stable isotopes to estimate trophic position: models, methods, and assumptions. Ecology 83, 703e718.

Rezende, C.E., Lacerda, L.D., Ovalle, A.R.C., Silva, C.A.R., Martinelli, L.A., 1990. Nature of POC transport in a mangrove ecosystem: a carbon stable isotopic study. Estuarine, Coastal and Shelf Science 30, 641e645. Riera, P., Richard, P., 1996. Isotopic determination of food sources of

Crassos-trea gigas along a trophic gradient in the estuarine bay of Marennes-Ole´ron. Estuarine, Coastal and Shelf Science 42, 347e360.

Rodelli, M.R., Gearing, J.N., Gearing, P.J., Marshall, N., Sasekumar, A., 1984. Stable isotope ratio as a tracer of mangrove carbon in Malaysian ecosys-tems. Oecologia 61, 326e333.

Round, F.E., Crawford, R.M., Mann, D.G., 1990. Diatoms. Cambridge Univer-sity Press, Cambridge, 747 pp.

Sobczak, W.V., Cloern, J.E., Jassby, A.D., Mu¨ller-Solger, A.B., 2002. Bio-availability of organic matter in a highly disturbed estuary: the role of de-trital and algal resources. Proceedings of the National Academy of Sciences of the United States of America 99, 8101e8105.

Stephenson, R.L., Tan, F.C., Mann, K.H., 1984. Stable carbon isotope variabil-ity in marine macrophytes and its implications for food web studies. Ma-rine Biology 81, 223e230.

Stribling, J.M., Cornwell, J.C., 1997. Identification of important primary pro-ducers in a Chesapeake Bay tidal creek system using stable isotopes of car-bon and sulfur. Estuaries 20, 77e85.

Su, H.M., Lin, H.J., Hung, J.J., 2004. Effects of tidal flushing on phytoplank-ton in a eutrophic tropical lagoon in Taiwan. Estuarine, Coastal and Shelf Science 61, 739e750.

Sullivan, M.J., Moncreiff, C.A., 1990. Edaphic algae are an important compo-nent of salt marsh food-webs: evidence from multiple stable isotope anal-yses. Marine Ecology Progress Series 62, 149e159.

Sweeting, C.J., Polunin, N.V.C., Jennings, S., 2006. Effects of chemical lipid extraction and arithmetic lipid correction on stable isotope ratios of fish tissues. Rapid Communications in Mass Spectrometry 20, 595e601.

Wainright, S.C., Weinstein, M.P., Able, K.W., Currin, C.A., 2000. Relative importance of benthic microalgae, phytoplankton, the detritus of smooth cordgrass Spartina alterniflora and common reed Phragmites australis to brackish-marsh food webs. Marine Ecology Progress Series 200, 77e91.

Ya´~nez-Arancibia, A., 1976. Observaciones sobreMugil curema Valenciennes, en areas naturales de crianza, Mexico. Alimentacion, madurez, crecimento y relaciones ecologicas. Anales del Centro de Ciencias del Mar y Limnologia. Universidad Nacional Auto´noma de Me´xico 3, 93e124.

數據

Fig. 1. The Guandu mangroves, Tapong Bay and their surrounding land uses.
Fig. 2. Classification of the diet composition of individual fish of Liza macrolepis with different total lengths (in mm) collected from (a) the Guandu mangroves and (b) Tapong Bay in the dry (D) and wet (W) seasons.
Fig. 3. d 15 N vs. d 13 C of food sources and dominant detritivorous fish collected from the Guandu mangroves in the dry (a) and wet (b) seasons
Fig. 4. d 15 N vs. d 13 C of food sources and dominant detritivorous fish collected from Tapong Bay in the dry (a) and wet (b) seasons

參考文獻

相關文件

Although Taiwan stipulates explicit regulations governing the requirements for organic production process, certification management, and the penalties for organic agricultural

Consistent with the negative price of systematic volatility risk found by the option pricing studies, we see lower average raw returns, CAPM alphas, and FF-3 alphas with higher

Good Data Structure Needs Proper Accessing Algorithms: get, insert. rule of thumb for speed: often-get

We are importers in the textile trade and would like to get in touch with ______ of this line.(A)buyers (B)suppliers (C)customers

6 《中論·觀因緣品》,《佛藏要籍選刊》第 9 冊,上海古籍出版社 1994 年版,第 1

• To enhance teachers’ knowledge and understanding about the learning and teaching of grammar in context through the use of various e-learning resources in the primary

• Is the school able to make reference to different sources of assessment data and provide timely and effective feedback to students according to their performance in order

Wang, Solving pseudomonotone variational inequalities and pseudocon- vex optimization problems using the projection neural network, IEEE Transactions on Neural Networks 17