Resource Utilization Patterns of Epifauna from Mangrove Forests with Contrasting Inputs of Local Versus Imported Organic Matter

MARINE ECOLOGY PROGRESS SERIES Vol.278:77-88,2004 Published September 7 Mar Ecol Prog Ser

Resource utilization patterns of epifauna from mangrove forests with contrasting inputs of local versus imported organic matter

Steven Bouillon1,., Tom Moens2, Inge Overmeer1, Nico Koedam3, Frank Dehairs1

IDepartmentofAnalytical and Environmental Chemistry, Mangrove Management Group, Vrlje Unlversltelt Brussel, Plelnlaan 2, 1050 Brussels, Belgium 2Biology Department, Marine Biology Section, Ghent University, Krljgslaan 281/S8, 9000 Gent, Belgium 3Department of General Botany and Nature Management, Mangrove Management Group, Vrlje Unlversltelt Brussel, Plelnlaan 2, 1050 Brussels, Belgium

ABSTRACT: Mangrove epifaunal communities have access to various carbon and nitrogen sources and we hypothesized that the degree of material exchange with the aquatic environment might influence the overall use of different substrates by intertidal communities. Therefore, we analyzed C and N stable isotope ratios in primary producers, sediments and 245 samples of epifauna hand-collected from 5 sites in India, Sri Lanka and Kenya (representing estuarine, lagoonal and basin-type mangrove forests). Several patterns emerged from this data set. First, epifaunal communities used a range of available food substrates at all sites studied, including mangrove-derived organic matter, local microphytobenthos and micro-epiflora, as well as imported C and N from the aquatic environment (i.e. phytoplankton- and/or seagrass-derived organic matter). Secondly, our data indicate that at sites with significant inputs of aquatic sources, use of mangrove carbon is rather limited on a community basis, whereas in systems with less material exchange with adjacent waters, the relative importance of mangroves is higher. Thus, despite the unquestionable impact some epifaunal species may have on leaf litter dynamics, the dependency of the invertebrate community as a whole on mangrove litter is not ubiquitously large and varies according to the availability of local versus tidally imported sources. Precise quantification of the relative importance of different substrates with 013Cand 015Nis, however, not always straightforward due to the multitude of available sources and the overlap in source stable isotope signatures. Micro-epiflora on mangroves trees were remarkably depleted in 15N in all systems (015Nbetween -8.2 and -2.4 %0)and thus form an example where 015Nis a useful source indicator, as low 015N values of several gastropod species indicated substantial feeding on such epiflora.

KEY WORDS: Sesarmid crab. Mollusk. Intertidal. Invertebrate' Foodweb . Stable isotope' Carbon. Nitrogen

Resale or republication not permitted without written consent of the publisher

INTRODUCTION bon dynamics in mangrove and adjacent ecosystems are still far from understood (e.g. Bouillon et al. 2004), Mangrove forests are often a prominent feature of and in view of the various settings in which mangrove the coastline in the tropics and subtropics, and are forests may occur, it is imperative to obtain data from a increasingly being exploited by humans. A thorough range of contrasting mangrove ecosystems rather than understanding of the ecological functioning and of generalizing conclusions from any particular site (e.g. their interactions with adjacent ecosystems is impor- Ewel et al. 1998). tant in providing a scientific basis for the management Invertebrate communities in intertidal mangrove and/or restoration of mangroves. Many aspects of car- forests have access to a variety of carbon and nitrogen

.Email: [email protected] ~ Inter-Research 2004 . www.int-res.com 78 Mar Ecol Prog Ser 278: 77-88, 2004

sources: local inputs from mangroves as litterfall or as sumers results in little fractionation in the case of 13C part of the sediment organic matter pool, microphyto- (Le. consumer 013Cvalues are close to those of their benthos, a variety of epiflora and tidally imported diet, with a slight enrichment typically between 0 and sources such as phytoplankton- or seagrass-derived 20/00),and somewhat higher for 15N (a value of -3%0 is organic matter. The removal of large amounts of leaf often cited, but the actual degree of fractionation litter by crabs observed in some systems (see Lee 1998) varies as a function of taxonomy, food quality and envi- and the fact that some other well-studied mangrove ronmental factors, see e.g. Vanderklift &Ponsard 2003 invertebrates (e.g. the gastropod Terebralia palustris) for a recent review). have been observed to feed extensively on mangrove We hypothesized that the dependency of the epi- leaves (e.g. Fratini et al. 2000) has nourished the still faunal community on mangrove-derived organic mat- prevalent view that 'the great majority of the man- ter would vary across different environmental settings grove macrobenthos relies directly on the high produc- and that such differences would be related to the relation of the mangroves themselves, consuming either tive availability of different potential sources. There- leaf litter or detritus composed of decaying leaves' (e.g. fore, we compared data on resource utilization pat- Fratini et al. 2000, Ashton 2002). The observation that terns (as evident from 013C and 015N analyses) of a a significant proportion of leaf production is efficiently variety of epifaunal species from a number of contrast- removed and/or consumed by some epifaunal species ing mangrove sites (Table 1): (1) a seaward exposed (e.g. 6lafsson et al. 2002) in some mangrove systems, is site and a more inland protected site with adjacent sea- not necessarily contradictory to the idea that mangrove grass beds in Gazi Bay (Kenya), (2) 1 site in the epifaunal corrununities as a whole may not depend pri- Coringa Wildlife Sanctuary (CWS), a large estuarine marily on mangrove production as a direct food source. mangrove system without nearby seagrass beds in the Recent results from stable isotope studies (e.g. Chris- Godavari delta (Andhra Pradesh, India), and (3) 2 small tensen et al. 2001, Bouillon et al. 2002, Hsieh et al. mangrove sites along the coast of Sri Lanka: the basin 2002) indeed suggest that only a limited number of forest of Galle and the lagoonal forest in Pambala. species may rely substantially on mangrove carbon Whereas the sites in India and Kenya experience and that a range of other carbon sources are used by medium or high tidal amplitudes (spring tidal range of the invertebrate community. Nevertheless, the degree -2 and 3.2 m, respectively), the sites in Sri Lanka are to which such results can be generalized is not known. both characterized by a low tidal amplitude (typically In particular, mangroves occur in highly variable geo- ~0.20 m). We supplemented these data sets with a morphological settings, which greatly influence the compilation of relevant literature data in an attempt to availability of potential C and N sources: where tidal look for general patterns in the origin of the carbon exchange is significant, marine/estuarine sources such assimilated by different epifaunal groups across as phytoplankton- or seagrass-derived organic matter different mangrove ecosystems. can be imported, but in low tidal amplitude settings, mangroves are the dominant source of carbon available in the sedimentary pool (Bouillon et al. 2003). MATERIALS AND METHODS The use of stable isotopes as tracers of the origin and assimilation of organic matter by faunal communities Description of study areas. Gazi Bay, Kenya: Gazi has become widespread, and is based on the assump- Bay (39° 30' E, 4° 22' S) is a shallow, tropical coastal tions that (1) different sources (may) have different water system, located -50 km south of Mombasa 013Cand 015Nsignatures, and (2) assimilation by con- (Fig. 1a,b). The total area of the bay is approximately

Table 1. Overview of the main characteristics of the study sites. CWS: Coringa Wildlife Sanctuary, Andhra Pradesh

Site Forest type Dominant vegetation Tidal Adjacent seagrass or setting range (m) beds? Gazi Bay (Kenya) Upstream site Estuarine Mixed: Rhizophora mucronata, -3.2 Yes Avicennia marina, Ceriops tagal Seaward site Estuarine Dominated by Sonneratia alba -3.2 Yes CWS (India) Estuarine Mixed: Avicennia officinalis, A. alba, -2 No Excoecaria agallocha Galle (Sri Lanka) Basin Dominated by Rhizophora apiculata SO.20 No Pambala (Sri Lanka) Lagoonal (fringe) Dominated by R. apiculata SO.20 Yes

- Bouillon et al.: Resource utilization patterns of mangrove epifauna 79

Indian Ocean Ikm t

Fig. 1. Geographical location of the different sampling sites. (a) Overview, and location of (b) the 2 sampling sites in Gazi Bay, Kenya, (c) the sampling areas in the Coringa Wildlife Sanctuary, India, and (d,e) the sampling sites in Galle and Pambala, Sri Lanka, respectively. Note that the black square in (c) indicates the areas where similar data were collected and presented in Bouillon et al. (2002). The darkest areas represent the main mangrove-covered regions and the dotted area in (b) shows the location of the main seagrass beds. See Table 1 for an overview of the main site characteristics

10 km2, with an additional 6 to 7 km2 covered by Coringa Wildlife Sanctuary, Andhra Pradesh, India: mangroves, mostly Rhizophora mucronata, Sonneratia The Coringa Wildlife Sanctuary (CWS, between 82° 15' alba, Cenops tagal, Bruguiera gymnorrhiza, Avicennia and 82° 22' E, 16°43' and 17°00' N, Fig. la,c) is part of manna and Xylocarpus granatum. The bay itself har- the mangrove-covered area between Kakinada bay bors large areas (-70% of the total area) of often dense and the Gautarni branch of the Godavari river seagrass beds, which are dominated by Thalassoden- (Fig. la,c). The CWS is dominated by mangrove forests dron dliatum (Coppejans et al. 1992). The bay is open (covering -150 km2) and tidal mudflats, the most abun- to the Indian Ocean through a relatively wide and dant mangrove species being Avicennia marina, Avi- shallow entrance in the south. Besides a few tidal cennia offidnalis, Excoecaria agallocha, Sonneratia creeks without freshwater inputs, the upper region of apetala, Rhizophora mucronata and Rhizophora apicu- the bay receives freshwater from the Kidogoweni river, lata (Satyanarayana et al. 2002). Seagrass beds are which cuts through the mangroves. Spring tidal range absent from the areas adjacent to the mangroves. in Gazi Bay is reported to be 3.2 m (Kitheka 1997). Two Tides are semidiurnal and spring tidal amplitude in the intertidal sites were selected for the sampling of flora, bay is -2 m. Samples of vegetation, sediments and epi- sediments and invertebrates in July 2003: (Site 1) a fauna were collected in May and June 2001 in the mixed forest site located upstream along Kidogoweni northwestern part of the CWS, in an intertidal area creek, where vegetation consisted of R. mucronata, A. dominated by A. officinalis, A. marina and E. agallocha manna, C. tagal and a few X. granatum, and (Site 2) a along Matlapalem creek (Fig. la,c). seaward site predominantly vegetated by S. alba with Galle, Sri Lanka: The mangroves in Galle, southwest some R. mucronata and A. marina, which bordered the Sri Lanka (06° 01' N to 80° 14' E) cover an area of about edge of the bay and is, therefore, more likely to be 1.5 km2 and can be classified as a basin forest (sensu influenced by seagrass inputs (Fig. 1). Lugo & Snedaker 1974) (Fig. Ie). The forest is located

- - - - - 80 Mar Ecol Prog Ser 278: 77-88, 2004

approximately 600 m from the Indian Table 2. Overview of stable isotope data (average :I:1 SD) for flora, sediments Ocean shoreline and is traversed by 2 and epifauna from Gazi Bay, Kenya (Sites 1 and 2, see Fig. 1b) rivers, the Thalpe Ela and its tributary, the Galu Ganga. The forest at Galle is n further characterized by a very irregular topography due to the burrowing activity Site 1 (upstream site) of the mud lobster Thalassina anomala, Flora A vicennia marina -31.2:1: 0.9 3.0 :1:0.8 4 and large patches of the forest are per- Ceriops tagal -29.4 :I:0.9 0.1:1: 1.3 4 manently inundated. This site experi- Xylocarpus granatum -27.7:1:0.4 -0.7 :I:0.6 4 ences very low to no tidal influence and Rhizophora mucronata -29.0 :I:0.6 0.0:1: 0.7 4 hence, limited exchange of organic mate- Microphytobenthos -22.1 / -22.1 1.8 / 1.9 2 -24.2 -2.4 rial with adjacent environments, except Micro-epiflora 1 (pooled) during exceptionally high water levels in Sediments Surface sediments -25.2:1: 0.0 2.1 :I:1.2 3 the river. The majority of samples for All sediment layers, up to 10 em -25.2:1: 0.2 2.7 :1:0.9 15 Galle were collected in an area predomi- Mollusks nantly vegetated by Rhizophora apicu- Cerithidea decollata -21.6:1: 0.8 4.7 :I: 1.2 7 lata in March 2002 (dry season). Some Isognomon ephippium -22.8 :I:1.0 4.8:1: 0.3 5 initial samples were collected earlier in Littoraria scabra -25.2 :I:1.3 1.7:1: 0.3 3 February-March 2000; and some stable Trapezium efr. sublaevigatum -23.0:1: 0.2 4.9:1: 0.1 3 isotope data on vegetation and sediments Cassidula labrella -23.5:1: 0.6 5.6:1: 0.1 4 Onchidium spp. -23.7:1: 1.5 1.7 :I:0.6 5 were available from a previous study Saccostrea cucculata -23.7 4.8 1 (Bouillon et al. 2003). Terebralia palustris -26.1 :I:0.9 3.5 :I: 1.1 8 Pambala. Sri Lanka: Pambala-Chilaw Brachyuran crabs lagoon is situated in western Sri Lanka Epixanthus dentatus -22.6/-20.8 7.0/7.1 2 (07035' N, 79047' E) and is similarly char- Eurycarcinus natalensis -21.2/-21.4 6.6/7.2 2 acterized as a low tidal amplitude system Metopograpsus thukuhar -21.0/-21.0 6.9/7.2 2 5.5 1 (tidal range not exceeding 0.2 m, F. Dah- Neosarmatium smithi -30.3 Perisesarma guttatum -23.3 :I:0.9 4.2:1: 0.7 5 douh-Guebas pers. obs.), with -3.5 km2 Perisesarma spp. -22.1:1: 0.3 4.0:1: 0.3 5 of fringing mangroves (Fig. Id). Surface Uca chlorophthalmus -21.1 :I:0.7 3.2:1: 0.8 5 water salinity in the lagoon varies Uca urvillei -20.8 :I:0.8 4.1:1: 0.7 6 strongly, between 0 and 55. All samples Site 2 (seaward site) from Pambala were collected in a Rhi- Flora zophora mucronata-dominated area in Avicennia marina -30.4 :I: 1.3 3.3 :I:0.4 4 March 2002, but some previous data on Sonneratia alba -27.0 :I:0.8 2.2 :I: 1.0 4 4 vegetation and sediments were available Rhizophora mucronata -29.8 :I:0.8 -0.5 :I:2.2 (Bouillon et al. 2003). Sediments Surface sediments -23.0 :I:0.9 1.7:1: 0.8 3 In conclusion, the different sampling All sediment layers, up to 10 em -23.8 :I:0.9 2.1:1: 0.4 11 sites represent a variety of geomorpho- Mollusks logical settings (summarized in Table 1): Neritina spp. -17.9 4.2 1 (1) Pambala (Sri Lanka) as an example of Littoraria scabra -23.5 :I: 1.0 1.1 :I: 1.4 7 a low tidal amplitude lagoon with fring- Crassostrea cucculata -18.4 :I:0.6 5.1:1: 0.6 5 ing mangroves, (2) Galle (Sri Lanka) Terebralia palustris -22.9 :I:0.8 4.5:1: 0.5 6 being a basin forest with insignificant Clypeomorus spp. -19.0:1: 0.7 5.1 :I:0.4 5 tidal influence, (3) the CWS (India) as an Brachyuran crabs example of a large estuarine mangrove Uca lactea annulipes -16.4:1: 0.3 2.2:1: 1.1 6 Uca urvillei 5 forest with significant tidal influence but -16.6:1: 1.2 3.0 :I: 1.0 Perisesarma guttatum (adults) -20.1 :I:0.5 4.9 :1:0.5 5 without seagrass inputs, and (4) 2 sites in Perisesarma guttatum (juveniles) -19.1 5.2 1 the smaller estuarine system of Gazi Bay Metopograpsus thukuhar -20.5 :I: 2.2 5.5:1: 0.2 3 (Kenya), 1 of which is located in the inte- Thalamita crenata -25.4 3.0 1 rior part of the forest, the other site near Sarmatium crassum -23.6 4.4 1 Macrophthalmus depressus -16.4 :I:0.2 2.6:1: 0.5 5 the forest fringe and, therefore, more exposed to inputs from the extensive sea- Miscellaneous taxa Tedania digitata -18.5 :I: 1.3 4.9 :I: 1.8 4 grass beds in the bay. The strongest tidal Barnacles (unidentified) -18.3:1: 0.7 7.2:1: 0.4 3 influence is experienced at Gazi Bay, Clibanarius spp. -22.8/-21.2 4.9/2.8 2 Kenya.

------Bouillon et al.: Resource utilization patterns of mangrove epifauna 81

Sampling and analytical techniques. At all sites, OlJC = Xsample- Xstandard X 103 samples of epifauna and flora were taken from a 20 m Xstandard by 20 m area in order to avoid spatial variations in producer stable isotope signatures (e.g. Fry et al. 2000). Although our sampling design does not, therefore, consider potential spatial and seasonal variations in RESULTS the foodweb structure within sites, this should not limit our conclusions (see 'Discussion'). Sampling efforts Sedimentary organic matter were concentrated on those intertidal species (mainly mollusks and brachyuran crabs) that were numerically Sediment at the upstream Gazi site had OlJC values abundant in the area and, therefore, likely to play an (-25.2 :I: 0.2%0) somewhat more enriched than man- important role in the overall processing of organic mat- grove leaf tissues and showed relatively high TOC/TN ter. Samples of vegetation and epifauna were collected ratios (16.0 :I: 1.3 atom, Fig. 2). At the seaward site, by hand, while benthic rnicroalgae were obtained by however, OlJC values of sediments were higher and gently scraping them off the sediment where they more variable, particularly in the surface layers (the formed a conspicuous layer; this was only the case in latter on average -23.0 :I:0.90/00).Sediments in Galle Gazi Bay, although previous data were available from and Pambala showed consistently low OlJC values the CWS. Mangrove leaf samples were picked from (-27.8:1: 1.1 and -27.5:1: 0.90/00,respectively), high con- different trees to avoid any bias in the resulting isotope centrations of organic carbon and high TOC/TN ratios signatures. Typically, only surface sediments were col- (Fig. 2). Sediments in the CWS showed the highest lected, but for the 2 sites in Gazi, sediment cores were OlJC values and had low TOC and TOC/TN ratios taken in the framework of another study and parti- (Fig. 2). tioned into 0-1, 1-2,2-4 and 4-10 cm layers. All faunal samples were kept in a cool box, transported to the field laboratory, washed and dried at 60°C for at least Primary producer stable isotope signatures 48 h. For some of the smaller crab species (e.g. Uca spp., smaller sesarrnids) the gut and intestinal system Different mangrove species from the 2 sites in Gazi were first removed and muscle tissue of the body was all showed typical OlJC signatures, with averages rang- used; for larger crab species, muscle tissue was taken ing between -27.0 and -31.20/00(Table 2, Fig. 3). The 3 from the chelae. For mollusks, tissues were removed mangrove species collected in the CWS had o\3C val- from their shell and analyzed whole. All samples were ues similar to those found previously in the area (see ground to a fine powder and sub-samples for OlJCsub- Bouillon et al. 2002), but 2 species had markedly sequently acidified with dilute (5%) HCI before ana- higher Ol5N values (Avicennia marina: +8.9%0 and lysis to remove carbonates (except those of mangrove Excoecaria agallocha: +8.8 :I:0.3%0, Table 3, Fig. 3). tissues). Sediment total organic carbon (TOC) and total Mangrove leaf tissues from Pambala all showed rather nitrogen (TN) were determined by com busting pre-weighed samples in a ThermoFinnigan Flash1112 elemental -20 analyzer. OlJC and Ol5N analysis of ews a b flora, fauna and sediments was per- ews formed with the aforementioned elemental analyzer, coupled to a Ther- t+Gazi,seaward + ~i.seaward moFinnigan delta + XL via a Conflo III ~i. upstream -+- Gazi, upstream -26 interface, with a typical reproduci- ~E 'i3 bility of :1:0.150/00for both O\JC and ~ -28 Ol5N. The samples from the CWS, however, were all measured on a dual -30 inlet Finnigan Mat delta E off-line o 4 8 12 16 20 24 28 32 6 10 14 18 22 26 30 after cryogenic purification of the C02 Sedimenl TOe (%) Sediment TOefTN or N2, with a reproducibility of 0.20/00 for both OlJC and Ol5N.All stable iso- Fig. 2. Relationship between (a) sedimentary total organic carbon content (TOC) tope ratios are expressed relative to and 1)t3cof sediment organic carbon, and (b) sediment TOC/TN (total nitrogen) ratios (atom) and 1)13Cof sediment organic carbon, for the different systems/sites the conventional standards (VPDB studied. Note that some additional data for Galle and Pambala, and all TOC limestone and atmospheric N2) as 0 and TOC/TN data for the CWS (Coringa Wildlife Sanctuary) were taken from values (0/00),defined as: Bouillon et al. (2003)

-- 82 Mar Ecol Prog Ser 278: 77-88. 2004

8 a 8 -<>- b 6 6 4 4 -0- 2 2 ~o~t 0 o fL m1 -2 . -2 + -4 -4 -6 -6 Gazi, upstream Gazi, seaward -8 -8

. 8 ! I~ 18 c \4 6 !r d 4 ! . f ~-l - ]() !3 2 I'.' I' :I 6 !! ~ J '"z o vD D~_~_2_~I -?=tlYi . 7.0 2 F , . -2 . 1 I ,I 1. .I + -2 -4 .I I. 1: : .I I. -6 T---m---- -6 .1 I. . CWS !. ! Galle -10 -.- -8 -35 -33 -3\ -29 -27 -25 -23 -21 -19 -17 -IS 8 I ! e /) 13 C (%0) 6 e ~ I , Mangrove tissues Sediments 4 o 6 ~D-4- 00 ,: ,: Micro-epillora . Mollusks 2 : : ie! J , I I!EIM icrophytobenthos . o ~.~ L J o Brachyuran crabs -Tl. o Miscellaneous epifauna -2 .I 1. :--1 Typical range for microphytobemhos -4 i ! 1___1 I. I. (literature data) -6 I. .I . ! ! Pambala sUe range for mangrove tissues -8 (data from Roggeman 2(02) -35 -33 -3] -29 -27 -25 -23 -2\ -]9 -17 -15 13 /) C (%0) Fig. 3. Plot of 1)15Nversus 1)13efor epifauna from the 5 study sites. (a) Upstream site along Kidogowenicreek (GaziBay,Kenya). (b) the seaward site (Gazi Bay. Kenya). (c) Coring a Wildlife Sanctuary, India. (d) Galle, Sri Lanka and (e) Pambala lagoon, Sri Lanka. Note the different scale of the y-axis in (c). See Table 1 for site characteristics low /)13Cvalues, with averages ranging between -29.2 and Pambala. and -2.4 '1'00in Gazi. 1)13Cvalues for these and -31.2'1'00for various mangrove species (Table 4, epiphytes were highly variable: -21.4'1'00in the CWS, Fig. 3). The most depleted /)t3C values for mangrove -24.2%0 in Gazi, -29.2'1'00in Galle and -32.0'1'00in Pam- tissues, consistently lower than -30%0, were found in bala (Tables 2 to 4, Fig. 3). It should be noted that all Rhizophora apiculata and Bruguiera gmnorrhiza from data of micro-epiflora were gathered on pooled sam- Galle (Table 4, Fig. 3). E. agallocha, however. which ples and we have no indications of the variability in mostly grows at slightly elevated patches in this area, isotope signatures of this source. Due to difficulties in had /)13Cvalues in the usual range. averaging -28.6'1'00. sampling, we only have data for microphytobenthos In contrast to mangrove tissues, micro-epiflora grow- from the upstream site in Gazi (1)13C:-22.0'1'00,1)15N: ing on mangrove stems (the composition of which was +1.8'1'00),but previous data from the CWS were avail- not studied) showed consistently low 1)15Nvalues in all able (1)13C: -17.3:t 1.7'1'00, 1)15N: +1.7:t 1.7'1'00, see Bouil- areas studied: -8.2 '1'00in the CWS, -6.8 '1'00in both Galle lon et al. 2002). Bouillon et al.: Resource utilization patterns of mangrove epifauna 83

Table 3. Overview of stable isotope data (average :I:1 SD) for flora, sediments highly variable both within and and epifauna from the Coringa Wildlife Sanctuary, Andhra Pradesh, India. Note between different systems (overall that additional data from this area (but different localities within the CWS) range: -30.3 to -18.90/00,with a fairly can be found in Bouillon et al. (2002) even distribution between these ex- tremes). In the ews, for example n (Table 3, Fig. 3), average 1)l3e values Flora for sesarmids were -25.4 0/00(adult Avicennia marina -29.0 8.9 1 (pooled) Perisesarma bengalensis and Epis- Avicennia officinalis -30.1:1: 0.6 3.5 :1:0.9 4 esarma versicolor), -24.0 (juvenile P. -26.6:1: 0.6 8.8:1: 0.3 5 Excoecaria agallocha bengalensis), -23.80/00 (Parasesarma Micro-epiflora -21.4 -8.2 1 (pooled) asperum), and -19.50/00 (juvenile Pa- Sediments rasesarma plicatum). As previously Surface sediment -24.9 5.6 1 (pooled) Mollusks noted for this area (Bouillon et al. Pythia plicata -25.3 :I:0.3 3.8 :I: 1.8 4 2002), ocypodid crabs (3 Uca spp.) Onchidium spp. -22.2:1: 0.8 -5.0 :I:2.6 4 and Metaplax distinctus were found Teridinidae spp. -24.6:1: 0.4 5.8 :I: 1.5 5 at the more enriched end of the 1)l3e -27.7 19.5 1 Polymesoda bengalensis range, and mollusks showed a diverse Brachyuran crabs 1)13e and 1)15N pattern (Table 3, Episesanna versicolor -25.4 :I:0.5 6.6:1: 2.1 4 Perisesarma bengalensis (adults) -25.4 :I:0.5 7.6:1:1.1 6 Fig. 3). With the exception of Uca Persisesarma bengalensis (juveniles) -24.0:1: 0.8 5.0:!: 0.7 6 lactea annulipes, invertebrates from Parasesanna asperum -23.8 :I:2.0 7.3:!: 0.7 4 Galle all showed relatively uniform 4 Parasesarma plicatum -19.5:1: 0.4 3.8:1: 0.8 1)13e values (Table 4, Fig. 3), with Metaplax distinctus -22.5/-22.9 7.0/7.0 2 Uca rosea -20.7 6.9 1 averages between -27.5 and -24.80/00. Uca triangularis -21.7/-21.7 7.8/7.8 2 However, 1)15Nvalues ranged more Uca urvillei -17.4:1: 1.1 7.6:!: 1.3 4 widely, with unusually low values for Pythia plicata (-2.4 :f:0.7%0)and high values for the 2 bivalves examined (i.e. Polymesoda spp. and an uniden- Consumer stable isotope signatures tified oyster, 5.7 :f:1.4 and 8.0 :f:0.20/00,Table 4, Fig. 3). Invertebrates from Pambala, with the exception of Invertebrates from both sites in Gazi showed quite Cerithidea cingulata, similarly showed a fairly narrow diverse stable isotope signatures, ranging overall range of 1)13e signatures, with averages between between -30.3 and -16.4%0 for 1)13e,and between 1.1 -28.5 and -24.30/00(Table 4, Fig. 3). As in Galle, the and 7.20/00for 1)15N(Table 2, Fig. 3). However, 1)13e filter-feeding Polymesoda spp. showed distinctly values were typically higher at the seaward site higher 1)15Nvalues (8.4 :f: 0.50/00)compared to other (Fig. 3). Our 1)13edata on mangrove sesarmids are invertebrates (0.6 to 6.20/00).

-18 -14 -18 a b . c -20 -16 -20 u -18 u +-t-.1 + +.,' /I~V 70 -22 "'u -20 70 -22 -0 * / "8 <0 -22 ,/ 0. -24 .~ -24 8 / g ::::.-24 ,;' ~ -26 . -26 -26 8 -28 -28 -28 -30 -30 -30 -30 -28 -26 -24 -22 -20 -18 -3D-28 -26 -24 -22 -20 -18 -16 -14 -30 -28 -26 -24 -22 -20 -18 Sediment Ol3C

Fig. 4. Relationship between surface sediment 513Cvalues and those recorded in various species of (a) Sesanninae, (b) Uca spp. and (c) gastropods (data from all species at a specific site were pooled for c). Data were compiled from the 5 sites mentioned in this study and additional literature data (data sources available on request). .: literature data; 0: data from this study. Grey shaded areas show typical range of values for mangrove-derived organic matter. Literature data taken from Fry (1984), Rodelli et al. (1984), France (1998), Thimdee et al. (2001), Bouillon et a1. (2002) and Hsieh etal.(2002)

------84 Mar Ecol Prog Ser 278: 77-88, 2004

Table 4. Overview of stable isotope data (average:!: 1 SD) for flora, sediments, reflect the balance between local inputs particulate organic carbon (POC) and epifauna from Pambala and Galle, from mangroves (high TOe, high Sri Lanka TOerrN and low 1)13C)and inputs from the water column (low TOe, 1)13C 1)15N n low TOerrN and higher but variable 1)13C).Thus, our data show that both the Pambala, Sri Lanka Flora fringing lagoonal mangroves in Pam- Avicennja offjdnaljs -29.9:!: 1.8 0.2 :!: 1.3 5 bala and the basin forest of Galle are Brugujera gymnorrhjza -31.1 :!: 1.0 2.1 :!: 1.5 4 'retention' sites, where mangrove car- Rhjzophora apiculata -29.2 :!: 1.0 4.0 :!:0.7 5 bon accumulates and dominates the Rhjzophora mucronata -31.2:!: 0.9 4.8:!: 1.5 15 sedimentary organic matter pool. For Micro-epiflora -32.0 -6.8 1 (pooled) Sediments the ews and Gazi, however, a balance Surface sediments -27.5:!: 0.9 -0.1 :!:0.3 18 (4 for 1)15N) exists between local mangrove inputs Mollusks and tidally imported organic matter Pythja plicata -26.4 :!:0.5 0.6 :!: 1.3 4 (see also Hemminga et al. 1994, Slim et Cassjdula mustelina 2.7/2.7 2 -28.0/-29.0 al. 1996). Moreover, the data from the Cerithjdea dngulata -21.3/-21.0 6.2 / 6.3 2 Polymesoda spp. -29.5 :!: 1.5 8.4 :!:0.5 3 seaward site in Gazi show that seagrass Brachyuran crabs inputs (with 1)13evalues in the area Epjsesarma tetragonum -25.2 3.2 1 between-19.7 and-10.70/00,Hemminga Perisesarma dussumierj -27.3 :!:0.8 4.2:!: 1.2 6 et al. 1994, authors' unpubl. data) are Pseudosesarma crassjmanum -24.3 3.5 1 more important at this location than at Miscellaneous Isopoda (unidentified) -25.8 :!:0.6 1.5 :!: 1.6 4 the upstream sampling site.

Galle, Sri Lanka Flora Rhjzophora apiculata -31.8:!: 1.3 2.2:!: 5.0 13 (7 for 1)15N) Consumption of different primary Excoecarja agallocha -28.6:!: 1.6 O.O:!: 0.7 8 (6 for 1)15N) food sources by mangrove epifauna Brugujera gymnorrhjza -33.7 :!:1.1 1.5 :!: 1.3 9 (7 for 1)15N) -29.2 Micro-epiflora -6.8 1 (pooled) Before discussing the differences in Sediments and POC e and N sources for epifaunal commu- Surface sediments -27.8:!: 1.2 1.8:!: 1.7 17 (15 for 1)15N) POC -26.9 :!:1.3 6.8:!: 2.6 6 nities between different sites, we will Mollusks first briefly examine the major primary Terebralia palustris -24.8:!: 1.7 4.8:!: 1.5 10 organic matter sources for the 2 domi- Cassjdula muste1jna -25.8 :!:0.6 3.7:!: 0.6 6 nant groups of epifauna, i.e. brachy- -2.4 :!:0.7 6 Pythja plicata -25.8 :!:0.6 uran crabs and mollusks, at the Polymesoda spp. -27.5 :!: 1.8 5.7 :!:1.4 6 different sites studied. Unidentified oyster -25.7 :!: 1.1 8.0:!: 0.2 5 Brachyuran crabs Perisesarma dussumieri -25.8 :!: 1.5 3.9 :!: 1.5 6 Perisesarma bengalensjs -25.7:!: 0.1 2.9:!: 0.7 3 Brachyuran crabs Uca lactea annuljpes -18.3 :!:0.7 4.8:!: 0.7 5 Miscellaneous Brachyuran crabs and mollusks are Clibanarius spp. -25.0 :!: 1.5 4.9:!: 1.0 5 typically the dominant groups of mangrove epifauna and can attain very high densities (e.g. Ashton & Macin- DISCUSSION tosh 2002). Their role in nutrient cycling (Lee 1998) and in sediment biogeochemistry, via bioturbation, is Differences in sedimentary organic matter origin therefore considered to be large. Even though the focus of most studies has been on the effect of sesarmid The different sites in this study represent different crabs on leaf litter dynamics (e.g. Ashton 2002), several geomorphological settings (i.e. estuarine, lagoonal and studies have stressed that different groups of man- basin forests, sensu Lugo &Snedaker 1974) with varying grove crabs display a wide range of feeding prefer- tidal amplitude and, therefore, the relative importance of ences. Although for several of the larger species 1)13e local vascular plant material and aquatic organic matter values were on average within -40/00of those of man- sources to the sedimentary pool varies significantly grove tissues (which indicates substantial inputs from (Fig. 2). It has previously been shown (Bouillon et al. mangrove litter, albeit with some contributions of other 2003) that for the majority of systems, these parameters dietary sources), several species showed consistently Bouillon et aI.: Resource utilization patterns of mangrove epifauna 85

high 1)13Cvalues tending towards those typical of reliance on mangrove carbon, as aquatic microalgal microphytobenthos (e.g. Perisesarma spp. from Gazi: production may be similarly depleted in 13Cdue to low -20.1 :I:0.5 and -22.1 :I:0.30/00,Parasesarma plicatum 1)13Cvalues in the dissolved inorganic carbon (DlC) from the CWS: -19.5 :I:0.4 0/(0).Thus, although some pool (e.g. surface water 1)13CDlCvalues in Galle ranged mangrove sesarmids undoubtedly rely substantially on between -14.2 and -7.80/00, authors' unpub1. data, mangrove carbon, this is clearly very species-specific. whereas typical seawater values are close to 0%0).The Secondly, when comparing sesarmid 1)13Cdata from all data in Table 4 (see also Fig. 3) also show that these the sites studied here and with the inclusion of litera- bivalves were consistently enriched in 15N compared ture data, we find (despite the species-specific varia- to most other consumers in the ecosystem, despite sim- tion) a good overall correlation between the isotope ilar 1)13Cvalues. For Galle, the only site with significant signature in sedimentary organic carbon and that in permanently inundated sites, suspended organic mat- sesarmids (Fig. 4a, Spearman rank correlation test: p = ter was similarly enriched in 15Ncompared to sediment 0.0057, R2 = 0.37). The latter confirms the idea that organic matter. Thus, the situation for these bivalve many sesarmids may spend more time feeding on the species remains somewhat unclear, with strong contri- sediment surface than actively searching for fallen butions from either mangrove-derived material (after leaves (Skov & Hartnoll2002) and indicates that the C microbial processing in the water column, which can and N source for this group of epifauna will partially result in higher 1)15Nvalues, e.g. see De Brabandere et depend on the geomorphological settings of the sys- al. 2002)or from aquatic primary production. A special tem, as the latter is a primary determinant of the origin group of bivalves are wood-boring species such as the of sedimentary organic matter (Bouillon et a1. 2003). Teredinidae (which were sampled in the CWS), known Such a correlation with sediment 1)13Cvalues is not sig- to harbor symbiotic cellulolytic bacteria capable of nificant (Spearman rank correlation test: p =0.349, R2 = N2-fixation. Our 1)13Cdata of Teredinidae (-24.6 :I: 0.06) for the second major group of brachyurans found 0.4 0/(0)are consistent with A vicennja wood (1)13C: in mangroves, ocypodid crabs (mostly Uca spp. and -25.70/(0)being the major C source, but the 1)15Nsigna- Macrophthalmus spp., see Tables 2 to 4). Ocypodid tures of the Teredinidae (+5.8 :I:0.5) are only slightly crabs are typical deposit feeders which forage on the higher than those of the log (1)15N:+5.10/(0)in which sediment surface. Stable isotope data presented earlier they were collected, indeed suggesting an additional (e.g. Rodelli et a1. 1984, Marguillier et a1. 1997, France input of N2-fixation to the N-requirements of these 1998, Hsieh et a1. 2002) and in this study consistently bivalves. Another bivalve from dead mangrove wood show high 1)13Cvalues, ranging between -24.5 and in Gazi, Trapezium efr. sublaevigatum (which does not -12.50/00,with an average of -18.9:1: 2.30/00,and with a have such a symbiotic relationship), showed 1)13Cand distribution markedly different from that of sesarmids. 1)15Nvalues markedly more distant from those of man- These data indicate a clear selectivity for l3C-enriched groves (Table 2, Fig. 3), suggesting additional inputs carbon sources such as microphytobenthos. Among from aquatic C and N sources. the other brachyuran crab taxa associated with man- Two previous studies have found unexpectedly low groves, Eurycardnus natalensis, Epixanthus dentatus 1)15Nvalues in some mangrove mollusks (Christensen and Metopograpsus thukuhar from Gazi Bay clustered et a1. 2001, Bouillon et al. 2002) and similar data were close together and their stable isotope signatures (high obtained here, i.e. for Pythia pljcata from Galle, 1)13Cand 1)15N,Table 2) are suggestive of little direct Onchidium spp. from the CWS and to a lesser extent inputs from mangroves and of significant predation on from Gazi, and for Littoraria scabra from Gazi (see lower trophic levels, consistent with literature data Tables 2 to 4, Fig. 3), all of which are species typically (e.g. Dahdouh-Guebas et a1. 1999). found grazing on mangrove roots or stems. The unusually low 1)15Nvalues found here in micro-epiflora (1)15N = -6.8'X>oin both Pambala and Galle, 1)15N= -8.20/00in Mollusks the CWS, -2.4 0/00in Gazi Bay) offer a convincing expla- nation for these consumer 1)15Nvalues and indicate a Most of the scarce, previously published stable iso- significant input of N from such a source. The 1)15Nval- tope data for bivalves from intertidal mangrove habi- ues of the epiflora are likely to reflect the importance tats (Rodelli et a1. 1984) point towards microalgal food of atmospheric nitrogen (from precipitation and/or sources, but the data presented here show a more through Nrfixation) as their N source (see e.g. Hietz et diverse pattern. Both the Polymesoda spp. (sampled in a1. 1999 for vascular epiphytes). It is worth noting that the CWS, Galle and Pambala) and the unidentified some other gastropods often found in large numbers oyster from Galle show consistently low 1)13Csigna- on mangrove stems, such as Cerithidea obtusa (found tures, close to those of mangrove tissues. However, this in the CWS, see Table 3) or C. decollata (found in Gazi, resemblance does not necessarily indicate a strong see Table 2) do not show evidence of extensive feeding

------86 Mar Ecol Prog Ser 278: 77-88, 2004

on 15N-depleted epiflora, but as these species are also mucronata leaves from both areas during the same sam- found on the sediment surface, the isotope data sug- pling period and found typically higher 813e values, gest that these are more likely to feed on sediment ranging between -29 and -250/00.A clear relationship organic matter and microphytobenthos. In conclusion, with tree age (as measured by the circumference of the the 813e and 815Npatterns (Fig. 3) indicate a diverse tree crown or the height of the tree) was found, with range of feeding preferences for mangrove-inhabiting younger trees showing more negative 813e values (be- mollusks. tween -32 and -30%0). A generally more enriched 813e signature for mangrove leaf tissues at both Galle and Pambala would be consistent with indications from sed- Differences in e and N utilization between imentary TOe and TOefTN data that the organic matter contrasting mangrove systems pool is derived primarily from mangrove tissues, and the sediment 813e signatures (-27.5 :I:0.9%0 in Pambala, In a previous study, we found no evidence for a dom- -27.8:1: 1.2%0in Galle) are indeed closer to the data from inant role of mangroves in sustaining epifaunal inver- Roggeman (2002). In this interpretation, the trophic tebrate communities (Bouillonet a1.2002).However, as dependency of epifauna on mangroves in both Galle and these data came from an estuarine system where Pambala would overall be large, and likely to occur both tidally imported organic matter was found to dominate directly through feeding on mangrove leaves and indi- the sediment pool. we hypothesized that such a situa- rectly through feeding on sediment organic matter tion may not be representative for other types of man- (which is strongly enriched in N compared to senescent grove systems, in particular for those where exchange mangrove leaves, e.g. see Skov &Hartno1l2002). Finally, of material with the aquatic environment is limited due it should be mentioned that our sampling design did not to low tidal action. accommodate potential seasonal and/or spatial varia- The data gathered in this study provide a first test of tions in the foodweb structure within sites. However, this hypothesis. First, as noted above, we found that even if such variations were present, we do not see any the origin of sedimentary organic matter at least par- arguments to conclude that these would interfere with tially determines the overall dependency of sesarmid the general patterns observed in Figs. 3 &4. crabs on different primary sources (Fig. 4a), despite the obvious fact that different species have different feeding specializations. In Fig. 4c, data for all gastropod Results of litter removal experiments and stable species are presented per site with data from this isotope analyses: how compatible are they? study, as well as those given in Rodelli et a1. (1984) and Bouillon et a1. (2002). Again, despite their varying The idea that much of the mangrove litter produced feeding specializations (some being selective for is removed and/or consumed by mangrove epifauna epiflora, microphytobenthos or mangrove litter, see (at least in the Indo-Pacific, McIvor & Smith 1995 and 'Mollusks' above), an analogous pattern as that found even there not in all forest zones alike or depending on for sesarmids shows up for gastropods, i.e. a markedly the tidal stage, see Slim et aI. 1997 and 61afsson et al. lower direct use of mangrove carbon in estuarine forest 2002) seems to contrast with the stable isotope results types where inputs from the aquatic environment are presented and compiled in this study. They demon- significant (Spearman rank correlation test for the rela- strate that, from a community perspective, only a lim- tionship between sediment and gastropod 813e: p = ited number of species rely substantially and directly 0.029, R2= 0.86). It is worth mentioning that the slope on mangrove carbon, and that, when available, a of the relationship between 813e of consumers and range of other sources are used by the invertebrate sediments is markedly different for sesarmids (slope community. There are several points which can be of -0.56) and gastropods (slope of -1.0). raised to reconcile these 2 superficially contrasting The same conclusions are also evident when compar- viewpoints. First, the fact that a large proportion of the ing the panels in Fig. 3: much of the species data for the leaf litter is removed and/or consumed by the local low-amplitude sites in Sri Lanka cluster quite closely crab fauna does not necessarily imply that mangrove together in the range expected for species that assimilate leaves are the dominant item in their diet, as the popu- significant mangrove-derived organic matter (from the lation of sesarmids may consume even more of 'some- sediment pool); for the upstream site in Gazi and for the thing else'. In this respect, the study of Skov &Hartnoll ews the data are more scattered; and at the seaward (2002) is particularly enlightening, as their field obser- site in Gazi, a clear trend towards much more enriched vations clearly demonstrate that sesarmids spend con- values is evident. However, for both Pambala and Galle, siderably more time feeding off the sediment surface our 813e values for most mangrove tree species are than collecting or eating leaves. Even though this does rather low. Roggeman (2002) analyzed Rhizophora not necessarily imply that more sediment organic mat-

------Bouillon et al.: Resource utilization patterns of mangrove epifauna 87

ter is assimilated than mangrove leaves, it does indi- Acknowledgements. Funding for this work was provided by cate the importance of deposit feeding. Evidently, the the Fonds voor Wetenschappelijk Onderzoek (FWO-Vlaan- sources of organic matter present in the sediment and deren, contract G.0118.02) and by EC-INCO project contract ERB IC18-CT98-0295. S.B. and T.M. are both postdoctoral fel- the degree of selectivity with which sesarmids feed on lows with the FWO- Vlaanderen. We are grateful to K. Ratnam it (both of which may be highly variable) will further for help during fieldwork in India, to 1.De Mesel, A. V. Borges determine which carbon sources contribute to their and the staff of the Kenya Marine and Fisheries Research diet and in which proportions. Secondly, much of the Institute (Mombasa) for fieldwork assistance in Kenya. to V. de Schuyter. M. Roggeman and A. Verheyden for help during work on the trophic significance of different sources sample collection in Sri Lanka, to P. Davie (Queensland in mangrove ecosystems has focussed on a limited Museum. Australia) and D. P. Gillikin for their expertise in number of invertebrate groups or species, notably identifying most of the crabs, and to A. Callea for his help in sesarmids and a disproportionate number of studies on identifying some of the mollusks from Kenya. The excellent Terebralia palustris (e.g. Slim et al. 1997). From a com- help of J. Bosire and J. Kairo in organizing logistics in Kenya was much appreciated. D. Gillikin and 3 anonymous referees munity perspective, however, this may severely bias provided very insightful and constructive comments to an our view of the importance of mangrove litter, as the earlier version of this manuscript. often diverse invertebrate community apparently dis- plays a wide variety of feeding specializations. LITERATURE CITED Lastly, it should also be noted that in view of the significant differences in elemental ratios between differ- Ashton EC (2002) Mangrove sesarmid crab feeding experiments ent food sources available to intertidal consumers (e.g. in peninsular Malaysia. J Exp Mar BioIEcol 273:97-119 mangrove leaves have a very low N content, micro- Ashton EC,Macintosh DJ (2002) Preliminary assessment of phytobenthos is much richer in N, etc.), the contribu- the plant diversity and community ecology of the Serna tan mangrove forest. Sarawak, Malaysia. Forest Ecol Manag tions of C and N from any dietary source are not nec- 166:111-129 essarily equal, but likely to be proportional to the C/N Bouillon S, Koedam N, Raman AV, Dehairs F (2002) Primary ratios of the substrates. The latter implies that the producers sustaining macro-invertebrate communities in dependency in invertebrate communities on man- intertidal mangrove forests. Oecologia 130:441-448 Bouillon S, Dahdouh-Guebas F. Rao AVVS, Koedam N, grove-derived N will generally be less than their Dehairs F (2003) Sources of organic carbon in mangrove dependency on mangrove-derived C. sediments: variability and some possible implications for ecosystem functioning. Hydrobiologia 495:33-39 Bouillon S. Moens T. Koedam N. Dahdouh-Guebas F. CONCLUSIONS Baeyens W, Dehairs F (2004) Variability in the origin of carbon substrates for bacterial communities in mangrove sediments. FEMS Microbiol EcoI49:171-179 Our data strongly suggest that where multiple C and Christensen JT. Sauriau PG, Richard P. Jensen PD (2001) Diet N sources are available, intertidal mangrove epifaunal in mangrove snails: preliminary data on gut contents and communities exploit all available food resources with stable isotope analysis. J Shellfish Res 20:423-426 Coppejans E, Beeckman H. De Wit M (1992) The sea grass and clear and consistent differences in utilization patterns associated macro algal vegetation of Gazi Bay (Kenya). between different taxa or species. Secondly, we have Hydrobiologia 247:59-75 shown a strong influence of the relative inputs of local Dahdouh-Guebas F. Giuggioli M, Oluoch A, Vannini M, Can- and tidally imported carbon sources (as reflected in nicci S (1999) Feeding habits of non-ocypodid crabs from sedimentary 813Cvalues) on the relative importance of two mangrove forests in Kenya. Bull Mar Sci 64:291-297 De Brabandere L, Dehairs F. Van Damme S. Brion N, Meire P, mangrove carbon to several groups of epifauna. This Daro N (2002) 015N and Ol3C dynamics of suspended variability shows strong similarities to the variable organic matter in freshwater and brackish waters of the contribution of different carbon sources to sedimen- Scheidt estuary. J Sea Res 48:1-15 tary microbial communities observed across various Ewel KC, Twilley RR, Ong JE (1998) Different kinds of man- mangrove sites (Bouillon et al. 2004). grove forests provide different goods and services. Glob Ecol Biogeogr Lett 7:83-94 Although 815Nwas a poor source indicator except for France R (1998) Estimation of the assimilation of mangrove micro-epiflora, which showed consistently low 815N detritus by fiddler crabs in Laguna Joyuda, Puerto Rico. values in all sites (-2.4 to -8.2 0/00),we can expect that the using dual stable isotopes. J Trop EcoI14:413-425 importance of mangroves as a N source for invertebrates Fratini S. Cannicci S, Vannini M (2000) Competition and interaction between Neosarmatium smithi (Crustacea: is less than for C, due to the low N content of mangrove- Grapsidae) and Terebralia palustris (Molluska: Gas- derived organic matter. Finally, it is worth pointing out tropoda) in a Kenyan mangrove. Mar Bioi 137:309-316 that the areal extent of 'closed' mangrove systems (i.e. in Fry B (1984) I3C/12C ratios and the trophic importance of which mangroves appear to provide most of the carbon algae in Florida Syringodium filiforme seagrass meadows. Mar BioI 79:11-19 fuelling epifaunal and microbial communities) is rather Fry B. Bern AL. Ross MS. Meeder JF (2000) 015Nstudies of limited on a global scale, as the largest mangrove- nitrogen use by the red mangrove. Rhizophora mangle L. covered areas are found in estuarine and deltaic systems. in south Florida. Estuar Coast Shelf Sci 50:291-296

-- 88 Mar Ecol Prog Ser 278: 77-88, 2004

Hemminga MA ,Slim FJ, Kazungu J, Ganssen GM, Nieuwen- Rodelli MR, Gearing IN, Gearing PJ, Marshall N, Sasekumar huize J, Kruyt NM (1994) Carbon outwelling from a man- A (1984) Stable isotope ratios as a tracer of mangrove grove forest with adjacent seagrass beds and coral reefs carbon in Malaysian ecosystems. Oecologia 61:326-333 (Gazi Bay, Kenya). Mar Ecol Prog Ser 106:291-301 Roggeman M (2002) Stable carbon isotope composition of Hietz P, Wanek W, Popp M (1999) Stable isotopic composition mangrove wood in relation to environmental factors in of carbon and nitrogen and nitrogen content in vascular a historical perspective (in Dutch). MSc thesis, Vrije epiphytes along an altitudinal transect. Plant Cell Environ Universiteit Brussel, Brussels 22:1435-1443 Satyanarayana B, Raman AV, Dehairs F, Kalavati C, Chan- Hsieh HL, Chen CP, Chen YG, Yang HH (2002) Diversity of dramohan P (2002) Mangrove floristic and zonation benthic organic matter flows through polychaetes and patterns of Coringa, Kakinada bay, East coast of India. crabs in a mangrove estuary: 813C and 834Ssignals. Mar Wetlands Ecol Manag 10:25-39 Ecol Prog Ser 227:145-155 Skov MW, Hartnoll RG (2002) Paradoxical selective feeding Kitheka JU (1997) Coastal tidally-driven drculation and the on a low-nutrient diet: Why do mangrove crabs eat leaves? role of water exchange in the linkage between tropical Oecologia 131:1-7 coastal ecosystems. Estuar Coast Shelf Sci 45:177-187 Slim FJ, Hemminga MA, Cocheret de la Moriniere E, van der Lee SY (1998) Ecological role of grapsid crabs in mangrove Velde G (1996) Tidal exchange of macrolitter between a ecosystems: a review. Mar Freshw Res 49:335-343 mangrove forest and adjacent seagrass beds (Gazi Bay, Lugo AE, Snedaker SC (1974) The ecology of mangroves. Kenya). Neth J Aquat EcoI30:119-128 Annu Rev Ecol Syst 5:39-64 Slim FJ, Hemminga MA, Ochieng C, Jannick NT, Cocheret Marguillier S, van der Velde G, Dehairs F, Hemminga MA, de la Moriniere E, van der Velde G (1997) Leaf litter Rajagopal S (1997) Trophic relationships in an interlinked removal by the snail Terebralia palustris (Linnea us) and mangrove-seagrass ecosystem as traced by 813Cand 815N. sesarmid crabs in an East African mangrove forest (Gazi Mar Ecol Prog Ser 151:115-121 Bay, Kenya). J Exp Mar BioI EcoI215:35-48 McIvor CC, Smith TJ III (1995) Differences in the crab fauna Thimdee W, Deein G, Sangrungruang C, Matsunaga K (2001) of mangrove areas at a Southwest Florida and a Northeast Stable carbon and nitrogen isotopes of mangrove crabs Australia location: implications for leaf litter processing. and their food sources in a mangrove-fringed estuary in Estuaries 18:591-597 Thailand. Benthos Res 56:73-80 61afsson E, Buchmayer S, Skov MW (2002) The east African Vanderklift MA, Ponsard S (2003) Sources of variation in decapod crab Neosarmatium meinerti (de Man) sweeps consumer-diet delta N-15 enrichment: a meta-analysis. mangrove floors clean of leaf litter. Ambio 31:569-573 Oecologia 136:169-182

Editorial responsibility: Otto Kinne (Editor), Submitted: March 23, 2004; Accepted: June 8, 2004 OldendorVLuh~ Germany Proofs received from author(s): September 2, 2004