lable at ScienceDirect

Estuarine, Coastal and Shelf Science 80 (2008) 261–268

Contents lists avai

Estuarine, Coastal and Shelf Science

journal homepage: www.elsevier .com/locate/ecss

d13C and d15N biogeographic trends in rocky intertidal communities along thecoast of South Africa: Evidence of strong environmental signatures

Jaclyn M. Hill*, Christopher D. McQuaidCoastal Research Group, Department of Zoology and Entomology, Rhodes University, PO Box 94, Grahamstown 6140, South Africa

a r t i c l e i n f o

Article history:Received 4 June 2008Accepted 9 August 2008Available online 16 August 2008

Keywords:stable isotopesd13Cd15Nbiogeographyfilter feedersdiet

* Corresponding author.E-mail addresses: jaclyn.m.hill@gmail.com, j.hill@r

ru.ac.za (C.D. McQuaid).

0272-7714/$ – see front matter � 2008 Elsevier Ltd.doi:10.1016/j.ecss.2008.08.005

a b s t r a c t

Ecosystem dynamics driven by top-down controls have been well documented in rocky intertidalcommunities, while the effects of bottom-up influences are comparatively poorly understood. Wehypothesized that large-scale signatures of the physical environment may be identifiable along the SouthAfrican coastline as it is subject to two very different current systems (Benguela and Agulhas Currents)that profoundly influence primary production and thus both food type and availability. Through stableisotope analysis, we examined biogeographic patterns in multiple trophic levels at four sites alonga 1400-km stretch of South African coastline and investigated the dietary role of macroalgal-derivedorganic carbon in rocky intertidal communities. The general positioning of trophic groups was compa-rable across all sites, with animals from the same trophic levels grouping together and with a d15Nfractionation of 1–2& between levels. The species found at all sites demonstrated east–west d15Nenrichment, presumably reflecting a biogeographic shift in nitrogen sources linked to upwelling on thewest coast. Filter-feeders gave particularly clear results. Using discriminant analysis, mussels could becategorized into four geographic groups based on carbon and nitrogen signatures: east coast, southeastcoast, south-west coast and west coast. Barnacles and polychaetes showed similar geographic groupingsto mussels, but with shifts in actual values (1& depletion in d13C and 3& enrichment in d15N relative tomussels). This suggests that fractionation varies between species within a trophic level.IsoSource models showed that Ulva sp. made large contributions to the diets of two microalgal grazers(Siphonaria capensis and Scutellastra granularis) and this dietary dependence increased when movingfrom west to east coast, along the shoreline. Additionally, IsoSource models determined that relative tophytoplankton, macroalgae accounted for upwards of 60% of suspended particulate matter sampled fromthe shore (SPM; d13C and d15N) at three out of four sites and linear mixing models showed over 40% (allsites) and 50% (three sites) contribution of nearshore d13C and d15N, respectively, to the diet of allsampled filter feeders, inferring heavy dependence on macroalgal carbon. Numerous processes influencethe stable isotope composition of algae, obscuring direct links between macroalgae and their consumers.In light of this, the clarity of the biogeographic patterns of filter feeders is remarkable and demonstratesa very strong signature of the physical environment in the intertidal community.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

Many studies have illustrated strong top-down controls (e.g.grazing and predation) in rocky intertidal communities (e.g. Mengeand Farrell, 1989; Power, 1992; Menge et al., 1999). In contrast,although recognized as a significant influence in terrestrial (e.g.Fretwell,1977; Marquis and Whelan,1994; Chen and Wise,1999) andfreshwater ecosystems (Carpenter et al.,1985, 2001), few papers haveexamined the role of bottom-up effects (nutrients and productivity)in marine rocky-shore ecology (e.g. Menge, 1992; Bustamante et al.,

u.ac.za (J.M. Hill), c.mcquaid@

All rights reserved.

1995; McQuaid and Lindsay, 2000; Menge, 2000; Wieters, 2005). Theregulation of community structure through bottom-up controlsextends beyond the influence of localized processes, and incorporatesthe effects of oceanography and coastal hydrography (Seitz and Lip-cius, 2001; Menge et al., 2003; Blanchette et al., 2006). Where thesebottom-up effects are strong enough they may permeate up throughhigher trophic levels in aquatic systems (Carpenter et al., 2001). Ifcouplingof nearshore hydrographyand intertidal communityecologyis indeed tight (Menge et al., 1999; Nielsen and Navarrete, 2004;Blanchette et al., 2006), then it should be detectable along coasts withstrongly contrasting environmental conditions. The South Africancoast is powerfully affected by the eutrophic Benguela (west coast)and the oligotrophic Agulhas (east-south coasts) current systems(McQuaid and Payne, 1998), implying biogeographic differences in

J.M. Hill, C.D. McQuaid / Estuarine, Coastal and Shelf Science 80 (2008) 261–268262

the quality of nutrient and food resources available to intertidalorganisms. We tested the hypothesis of regional differences in qual-itative environmental signatures along this coastline using stableisotope analysis.

Modeling energy flow through food chains is a step towardsunderstanding trophic relationships (Monterio et al., 1991) andcombined measurements of d13C and d15N ratios can provideinformation on both food web structure and source materials(Fredriksen, 2003). The application of stable isotope analysis hascontributed to the resolution of food webs in both benthic andpelagic systems. Examples include determining the primary diet oflong-finned pilot whales (Abend and Smith, 1997); quantifying theimportance of benthic phaeophtyes to the Antarctic peninsularfauna (Dunton, 2001); resolving trophic level interactions in anupwelling system in Galicia in NW Spain (Bode et al., 2003) andevaluating the dispersal of fish farming waste in the westernMediterranean (Sara et al., 2004).

A few studies have also applied this technique to intertidalcommunities. Bustamante and Branch (1996) investigated thetrophic connections between consumer species and the impor-tance of macroalgal detritus in a rocky intertidal food web on thewest coast of South Africa, suggesting a strong dependence on kelp-derived organic carbon, while Page and Lastra (2003) proposedheavy dependence of intertidal soft substratum bivalves on re-suspended microbenthos. Likewise Hill et al. (2006) inferred strongdependence of intertidal mussels on nearshore suspended partic-ulate matter (SPM), which contains a large component of detritusthat is presumably macroalgal in origin. They showed both carbonand nitrogen enrichment from east to west around the SouthAfrican coast and found that mussels could be grouped into fourgeographical regions based on their isotopic signatures. Our studyaimed (1) to investigate biogeographic variation in d13C and d15Nsignatures of rocky intertidal organisms to assess whether thebiogeographic trends described for mussels exist in other filter-feeders and/or other trophic groups and (2) to test for large-scaleenvironmental signatures by examining the reliance of intertidalcommunities on macroalgal-derived organic carbon at four sitescovering a wide biogeographic range around the South Africancoastline.

2. Methods

2.1. Sample collection

Nearshore SPM samples (x 3) were obtained from 5-l surfacewater samples taken from the shore (0 km). Samples (x 3) ofbenthic microalgae were obtained by scraping the rock surface witha razorblade, giving a signature for the overall microalgalcommunity regardless of species composition. Specimens (x 3) ofintertidal consumers from different trophic levels and triplicatesamples of the most abundant macroalgal species (Table 1) werealso collected. All samples were obtained from St. Helena Bay, CapeAgulhas, Port Alfred and Port Shepstone (Fig. 1) between February–March 2006. Not all these species occur throughout the coast. Theencrusting rhodophyte Ralfsia verrucosa was collected from gardensof the territorial limpet Scutellastra longicosta (x 3) in Port Alfred inSeptember 2006. Due to logistical constraints, sampling only tookplace at four sites along the coastline, however these sites werechosen based results from Hill et al. (2006), which sampled muchmore extensively along the same coastline.

2.2. Sample preparation

Water samples were filtered through pre-combusted (500 �C,6 h) GF/F Whatman� filters (0.45 mm pore size), using a vacuumpump (�4 cm Hg) and then oven-dried at 60 �C for 24 h.

Zooplankton and other large particles were manually removedunder a dissecting microscope at 16� magnification. Muscle tissuefrom each consumer was removed, rinsed in distilled water (dH2O)and oven dried (60 �C, 48 h). Muscle tissue was used in this study asit has low turnover rates (Gorokhova and Hansson, 1999) and istherefore representative of a time-integrated diet. As muscle tissuehas minimal lipid content (Tieszen et al., 1983; Chu et al., 2000), nolipid extractions were performed on any tissue. All macroalgaewere rinsed in dH20, visible epiphytes were removed and the algaewere oven dried (60 �C, 48 h).

2.3. Isotopic analysis

d13C and d15N signatures of all samples were determined usinga continuous flow Isotope Ratio Mass Spectrometer (IRMS), aftersample combustion in on-line Carlo–Erba preparation units at theUniversity of Cape Town, South Africa. Samples were run alongwith working standards that were used to calibrate the resultsagainst International Atomic Energy reference materials. Resultsare expressed in standard delta notation, dX¼ ([Rsample/Rstandard]� 1)1000, where X is the element in question and R is theratio of the heavy over the light isotope. Precision of replicatedeterminations for both carbon and nitrogen was �0.05&.

2.4. Data analysis

In separate analyses, d13C and d15N signatures of mussels (boththose sampled from Hill et al., 2006 and from this study) and otherfilter feeders were analyzed using k-means cluster analyses basedon Euclidian distances and validated using Discriminant FunctionAnalyses (DFA). All analyses were performed using Statistica v7(StatSoft Inc. 2004). A two-source linear mixing model (Bustamanteand Branch, 1996) was applied to establish the percentage contri-bution of organic carbon and nitrogen to filter feeder diet bynearshore SPM, using the offshore (10 km) SPM values determinedfor each site by Hill et al. (2006):

%Isotope [ ½ðdfilter feeder L doffshore SPM L IÞ=ðdnearshore SPM L doffshore SPMÞ�100

where I is the average fractionation of d13C or d15N per trophic leveland d is the isotopic ratio of the sample. This mixing modelassumed an average fractionation value of 1& for d13C (DeNiro andEpstein, 1978; Fry and Sherr, 1984; Peterson and Howarth, 1987),while a d15N fractionation of 1.7& was used for mussels (asreported by Hansson et al., 1997; Raikow and Hamilton, 2001;Moore and Suthers, 2005) and 2.2& for the polychaete Gunnareacapensis, and the barnacles Tetraclita serrata and Octomeris angulosa(secondary consumers; Dunton, 2001).

The IsoSource model described by Phillips and Gregg (2003) wasapplied on two separate occasions; firstly as a two isotope system(d13C and d15N) with three sources (phytoplankton, Ulva sp. andPorphyra capensis or Sargassum incisifolium) to quantify thecontribution of sampled macroalgae to nearshore SPM, usingoffshore SPM (10 km; Hill et al., 2006) as a measure of phyto-plankton. Secondly, as a two isotope system (d13C and d15N) withthree sources (microalgae, Ulva sp. and P. capensis or S. incisifolium)to determine the importance (in terms of organic d13C and d15N) ofmacro vs. microalgae in the diet of two grazers, Siphonaria capensisand Scutellastra granularis, at each site. Both models were calcu-lated using an increment of 1% and mass-based tolerance of 0.5&.An average fractionation of 1& for d13C (as reported by DeNiro andEpstein, 1978; Fry and Sherr, 1984; Peterson and Howarth, 1987)and 2.4& for d15N (as suggested for limpets by Dunton, 2001) wasused in this model.

Table 1List of species and trophic groups sampled for each site along the South African coastline

Taxon Trophic group St. Helena Bay Cape Agulhas Port Alfred Port Shepstone

SPM O O O OMicroalgae Producer O O O OUlva sp. Chlorophyta Producer O O O OSargassum incisifolium Phaeophyta Producer X O O OEcklonia maxima Phaeophyta Producer O X X XPorphyra capensis Rhodophyta Producer O O X XRalfsia verrucosa Encrusting Rhodophyta Producer X X O XMytilus galloprovincialis Bivalvia Filter feeder O X X XPerna perna Bivalvia Filter feeder X O O OGunnarea capensis Polychaeta Filter feeder O O O XOctomeris angulosa Cirripedia Filter feeder X O O OTetraclita serrata Cirripedia Filter feeder O O O OOxystele tigrina Gastropoda Microalgal grazer O O O XOxystele variegata Gastropoda Microalgal grazer O O O OScutellastra granularis Gastropoda Microalgal grazer O O O OHelcion pectunculus Gastropoda Microalgal grazer O O O XSiphonaria capensis Gastropoda Microalgal grazer O O O OSiphonaria serrata Gastropoda Microalgal grazer X O O OScutellastra longicosta Gastropoda Macroalgal grazer X X O XScutellastra cochlear Gastropoda Macroalgal grazer O O O OParechinus angulosa Echinoidea Macroalgal grazer O O O XStomopneustes variolaris Echinoidea Macroalgal grazer X X X OBurnupena lagenaria Gastropoda Scavenger O O O ONucella cingulata Gastropoda Predator O X X X

J.M. Hill, C.D. McQuaid / Estuarine, Coastal and Shelf Science 80 (2008) 261–268 263

IsoSource models reported all possible combinations of eachsource contribution, examined in small increments (e.g. 1%), thatsatisfied the isotopic mass balance in the mixing model (Phillipsand Gregg, 2003), however, as some of the source proportions werereported to make contributions across a wide range of proportionalvalues (e.g. contributions of Ulva sp. to the diet of S. capensis atSt. Helena Bay could be between 12 and 60%), the generated meansare presented.

3. Results

3.1. Trophic level fractionation

Each site along the coastline showed similar isotopic positioningof trophic groups, with all organisms remaining in the range of�9.0to�18.0& for d13C and 4.0–14.0& for d15N. Although no consistentcarbon fractionation was seen between lower trophic groups, anoverall trend of nitrogen enrichment on the scale of 1–2& wasobserved with increase in trophic level. While the pattern was not

Fig. 1. Map of southern Africa, showing sample sites in bold along the coastline,shaded area¼ continental shelf (�500 m). The four biogeographic regions as found byHill et al. (2006) are delineated by circles with dotted line borders.

perfect for all sites, producers and SPM tended to be the mostdepleted; followed by macroalgal grazers. Microalgal grazers andfilter feeders were comparatively enriched and scavengers were themost enriched in nitrogen at three out of four sites (Fig. 2). Therelative trophic positions of SPM and filter feeders were as antici-pated at some but not all sites and species, as were the relativepositions of SPM and macroalgae. Although microalgae were in thesame position at three sites, the only consistent trophic positioningacross all sites was the scavenging gastropod Burnupena lagenaria,which demonstrated predictable fractionation (1.0–2.0& d13C asreported by DeNiro and Epstein, 1978; Fry and Sherr, 1984; Petersonand Howarth, 1987, 3.0–4.0& d15N; as reported by Minagawa andWada, 1984; Post, 2002) consistent with a primary diet of filterfilters and/or microalgal grazers. Overall, values of standard devi-ations were larger for carbon signatures than for nitrogen.

IsoSource models determined that macroalgae accounted for60% or more of the organic matter (carbon and nitrogen) in near-shore SPM relative to phytoplankton at all sites except Port Shep-stone (Fig. 3), where the nitrogen signature of nearshore SPM wasmuch more depleted than at the other sites.

3.2. Biogeographic trends

Of the six species common to all four sites, most were groupedgeographically into south-west/west coasts vs. south-east/eastcoasts (Fig. 4). This effect was clearer for nitrogen than for carbon.An overall pattern of enriched d15N on the south-west and westcoasts (Cape Agulhas, St. Helena Bay) and depleted d15N on thesouth-east and east coasts (Port Alfred, Port Shepstone) indicatesa biogeographic effect on nitrogen sources (Fig. 4). In the case ofd13C, separation of east/west sites was only clear for three of the sixspecies: the barnacle T. serrata, the gastropod Oxystele variegata andthe gardening limpet Scutellastra cochlear. For T. serrata andS. cochlear, specimens from the west coast were depleted relative tothe east, but for O. variegata the reverse was true.

No biogeographic trends were seen in d13C among any trophicgroup as a whole, but the addition of our mussel data to those ofHill et al. (2006) established that, with the exception of St. HelenaBay, isotopic signatures of mussel tissue confirmed the biogeo-graphic trends (Fig. 5A) reported by Hill et al. (2006). Although the

-20 -18 -16 -14 -12 -10 -8 -6 -4 -20

2

4

6

8

10

12

14

16Port ShepstonePort Alfred

-20 -18 -16 -14 -12 -10 -8 -6 -4 -20

2

4

6

8

10

12

14

16

RVSL

-20 -18 -16 -14 -12 -10 -8 -6 -4 -20

2

4

6

8

10

12

14

16Cape AgulhasSt. Helena Bay

-20 -18 -16 -14 -12 -10 -8 -6 -4 -20

2

4

6

8

10

12

14

16

δ15N

δ15N

δ15N

δ15N

δ13C δ13

C

δ13C δ13

C

SPMMicroalgaeMacroalgaeFilter Feeders

Microalgal GrazersMacroalgal GrazersScavengersPredators

Fig. 2. d13C and d15N signatures of all organisms sampled from each site along the coastline, broken down into trophic groups. SL¼ Scutellastra longicosta, RV¼ Ralfsia verrucosa.Values are means� SD.

J.M. Hill, C.D. McQuaid / Estuarine, Coastal and Shelf Science 80 (2008) 261–268264

other filter feeders sampled in this study (the polychaeteG. capensis, and the barnacles T. serrata and O. angulosa) did not fallinto the groups previously identified in Hill et al. (2006), theyshowed a similar biogeographic pattern, with the sole exception of

Fig. 3. Proportional contribution of sampled algae and phytoplankton to nearshoreSPM determined by IsoSource modeling (Phillips and Gregg, 2003).

G. capensis from Port Alfred. k-Means cluster analysis classifiedthem into three isotopic groups that represented animals collectedfrom the east, south-east, and south-west coasts (Fig. 5B). Subse-quent Discriminant Function Analysis showed that these clusterscould be identified by their isotope signatures alone with 93–100%accuracy. Although the biogeographic pattern for these filterfeeders was the same as for mussels (this study, Hill et al., 2006),the actual values were displaced. Signatures of barnacles andpolychaetes were on average depleted by 1& in d13C and enrichedby 3& in d15N relative to mussels from the same biogeographicregions (cf Fig. 5A and B). The two source linear mixing modelillustrated over 40% and 50% dependence on nearshore organiccarbon and nitrogen, respectively, for filter feeder diet (musselsincluded) at all sites (with the exception of all filter feeders at PortShepstone for d15N only; Fig. 6A and B).

IsoSource models also indicated that at every site, the macroalgaUlva sp. was the most important contributor to the diet ofS. capensis and S. granularis, both which are considered to includemicroalgae in their diets. Dependence on Ulva sp. increased whenmoving east along the coastline (with the exception of S. granularisat Port Shepstone; Fig. 7).

4. Discussion

The general positioning of trophic groups was comparableacross all sites, with animals from the same trophic levels groupingtogether and with a d15N fractionation of 1–2& between levels. Thespecies found at all sites demonstrated east–west d15N enrichment,and mussels and other filter feeders could be categorized into fourgeographic groups based on their carbon and nitrogen signatures.IsoSource models showed that Ulva sp. made large contributions tothe diets of two microalgal grazers (S. capensis and S. granularis)

Fig. 4. d13C and d15N signatures of the six organisms found at every site (C¼ St. Helena Bay, -¼ Cape Agulhas, B¼ Port Alfred, ,¼ Port Shepstone). Values are means� SD.

J.M. Hill, C.D. McQuaid / Estuarine, Coastal and Shelf Science 80 (2008) 261–268 265

and this dietary dependence increased when moving from west toeast coast, along the shoreline. Additionally, IsoSource modelsdetermined that in general, relative to phytoplankton, macroalgaeaccounted for upwards of 60% of suspended particulate mattersampled from the shore (SPM; d13C and d15N) and linear mixingmodels showed over 40% contribution of nearshore SPM (d13C andd15N) to the diet of all sampled filter feeders, inferring heavydependence on macroalgal carbon.

4.1. Photosynthetic isotope fractionation

No consistent d13C fractionation or clear dietary links in carbonwere observed between producers and primary consumers at anysite. The lack of consistent fractionation is not surprising. Not onlycould the main diet of intertidal consumers be something otherthan the sampled producers, or a mixture of multiple macroalgaein unknown proportions, but the isotopic signatures of primary

producers (including macro/microalgae and phytoplankton) aresubject to large variability due to isotope fractionation duringphotosynthesis. Fractionation of carbon isotopes during photo-synthesis can be influenced by changes in growth rates (Burkhardtet al., 1999a,b), light intensity (Descolas-Gros and Fontugne, 1985;Thompson and Calvert, 1994; Burkhardt et al., 1999b), or temper-ature (Wong and Sackett, 1978). Other factors such as carbonconcentrating mechanisms within macroalgae (Sharkey and Berry,1985) and the primary source (HCO3

� or CO2) of organic carbon(Laws et al., 1997) during photosynthesis may also alter isotoperatios within a single plant, community and/or species. As photo-synthetic fractionation may result in d13C variability on a scale of>10.0&, obtaining consistent and/or characteristic signatures fordifferent algal species is difficult, and consequently it is unrealisticto expect clear links between producers and primary consumers.This complicates assessments of macroalgal dependence at highertrophic levels.

Fig. 5. (A) d13C and d15N signatures of intertidal mussels including samples from thisstudy (Feb–Mar 2006) and those from Hill et al. (2006), groupings according to k-means cluster analysis (see Hill et al., 2006): (a) east coast, (b) south-east coast, (c)south-west coast, (d) west coast. (B) d13C and d15N signatures of intertidal filter feedersfrom this study (excluding mussels) groupings according to k-means cluster analysis:(i) east coast, (ii) south-east coast, (iii) south-west coast.

Fig. 6. Proportional d13C (A) and d15N (B) contribution of nearshore SPM to filter feederdiet.

J.M. Hill, C.D. McQuaid / Estuarine, Coastal and Shelf Science 80 (2008) 261–268266

4.2. Carbon and nitrogen trends in trophic groups

The general isotopic positioning of trophic groups was compa-rable across all sites, with animals from the same trophic levelgrouping together. The only exceptions to this overall trend werethe producers and macroalgal grazers. Variation in the signatures ofmacroalgae may be explained by isotope changes during photo-synthetic fractionation, while for their grazers, assimilation ofunique species (or proportions) of macroalgae would lead to highvariability. The d15N signatures showed trophic level enrichment,but with less fractionation (1.0–2.0&) than predicted (3.0–4.0& asreported by Minagawa and Wada, 1984; Post, 2002). Bustamanteand Branch (1996) found similar isotopic positioning amongtrophic groups on the west coast of South Africa, though many oftheir d15N signatures were significantly more depleted than any inthe present study.

The only consumer to demonstrate predicted fractionationvalues at all sites was the scavenging whelk B. lagenaria which,typically for a secondary consumer, exhibited less variation inisotope signatures than the primary producers and consumers(O’Reilly et al., 2002). Data from all four sites suggest that theprimary diet of B. lagenaria consists of filter feeders and/ormicroalgal grazers (Fig. 2). Conversely, although there is well

documented evidence that adults of the limpet S. longicosta grazeprimarily on the encrusting red alga R. verrucosa, which it gardens(Branch, 1971; McQuaid and Froneman, 1993; Lasiak, 2006),samples from Port Alfred showed S. longicosta to be more depletedin carbon than R. verrucosa and to be relatively enriched by only1.0& in d15N. The lack of predicted fractionation between S. long-icosta and R. verrucosa may be a consequence of algal isotopevariability but suggests a greater dependence of S. longicosta onother primary producers than previously recognized.

IsoSource modeling (Phillips and Gregg, 2003) showed that,with one exception, macroalgae accounted for more than 60% andphytoplankton less than 40% of the organic carbon and nitrogenfound in nearshore SPM at all sites (Fig. 3). Although there arenumerous other macroalgae in both the intertidal and subtidalenvironments that may influence the isotopic signature of SPM,these findings suggest that overall, phytoplankton makes minimalcontributions to nearshore SPM in terms of either carbon ornitrogen.

4.3. Biogeographic trends

The six organisms found at all four sites demonstrated overalld15N enrichment from east to west, and this pattern was also foundfor a pelagic top predator (the Cape Gannet) in a separate study inthe same region (Jaquemet and McQuaid, unpublished). Thisphenomenon presumably represents a geographic shift in nitrogensources moving from the oligotrophic east coast, dominated by theAgulhas Current, to the eutrophic, upwelling dominated Benguelasystem of the west coast. Documented isotopic gradients betweenoligotrophic and eutrophic conditions (Saino and Hattori, 1980;Minagawa and Wada, 1984) appear to reflect reliance on recyclednitrogen (especially ammonia) in oligotrophic waters, which isdepleted in d15N relative to the upwelled nitrate used in eutrophicsystems (Miyake and Wada, 1967). Biogeographic trends were lessclear for d13C. Although three of the six species found at all sites also

Siphonaria capensis

0.0

0.2

0.4

0.6

0.8

1.0

Scutellastra granularis

0.0

0.2

0.4

0.6

0.8

1.0

Sites

St. Hele

na Bay

Cape A

gulha

s

Port Alfre

d

Port She

pston

e

Pro

po

rtio

na

l c

on

trib

utio

n to

N

ea

rs

ho

re

S

PM

MicroalgaeUlva sp.

Porphyra capensis

Sargassum incisifolium

Ecklonia maxima

Fig. 7. Proportional d13C and d15N micro and macroalgal contribution to the diets ofSiphonaria capensis and Scutellastra granularis at all sites as determined by IsoSourcemodeling (Phillips and Gregg, 2003).

J.M. Hill, C.D. McQuaid / Estuarine, Coastal and Shelf Science 80 (2008) 261–268 267

demonstrated crude d13C separation between east and west coasts,the filter feeder T. serrata and the macroalgal grazer S. cochleardemonstrated west coast enrichments in d13C, while the microalgalgrazer O. variegata showed west coast depletion. The reasons forthis are unclear, but again may reflect differential fractionationbetween macroalgae (including macroalgal detritus) and micro-algae during photosynthesis. No biogeographic patterns wereevident among the remaining species or trophic groupings.

With the exception of St. Helena Bay, isotopic signatures ofmussel tissue sampled in this study conformed to the biogeo-graphic pattern of mussel signatures seen in Hill et al. (2006)(Fig. 5A). A similar pattern, but with a shift in both carbon (1&

depletion) and nitrogen (3& enrichment) emerged in theremaining filter feeders (Fig. 5B), suggesting that fractionationwithin trophic levels differs significantly among species. Accordingto Herendeen (2004) strong bottom-up effects frequently requirelow levels of intratrophic interference and heavy prey dependenceat the target level. The low level of interference (i.e. intratrophicpredation) in mussel and barnacle beds and their strong depen-dence on water column nutrition may explain why biogeographic

patterns are apparent in sedentary filter feeders and not in othertrophic levels (grazers, scavengers etc.). The application of a two-source linear mixing model (Bustamante and Branch, 1996) showedover 40% contribution of nearshore organic carbon to the diet of allsampled filter feeders at all sites, inferring a strong dependence onnearshore rather than offshore SPM. In contrast, nearshore SPMcontributed upwards of 50% organic nitrogen to all filter feeders atonly three out of four sites, suggesting a separate source of nitrogenat Port Shepstone (Fig. 6A and B). These biogeographic and trophiclevel gradients imply that significant bottom-up effects may regu-late processes further up the food chain in rocky intertidalecosystems.

The two limpet species we identified as microalgal grazers(S. capensis and S. granularis) have diets that include the sporelingsof macroalgae, and mature plants, as well as benthic microalgae(Branch, 1971). The importance (in terms of organic d13C and d15N)of macro vs. microalgae in their diets was determined using Iso-Source modeling (Phillips and Gregg, 2003) and revealed a cleartrend of increasing dependence on Ulva sp. for S. capensis whenmoving from west to east (Fig. 7). This suggests a biogeographicshift in its primary food sources between St. Helena Bay and CapeAgulhas. Scutellastra granularis showed a similar biogeographicincrease in Ulva sp. dependence from west to east, but with a sharpdecrease at Port Shepstone (Fig. 7). Branch (1971) suggested thatS. granularis may feed on P. capensis, Gelidium sp. and Ulva sp. inaddition to microalgae. The present study revealed that generallyUlva sp. made the greatest contribution to the organic content ofthe diets of these two species, indicating that macroalgae, includingspores and emerging germlings are an integral part of the diet ofthese two grazers. The minimal contribution of P. capensis to thediet of either limpet may be attributable to its strong seasonality(McQuaid, 1985) and the paucity of a signal characteristic ofSargassum heterophyllum can be explained in a spatial context, asS. heterophyllum is found much further down on the rocky shore,and is usually confined to mid-intertidal rock pools (De Clerck et al.,2005).

5. Conclusions

d15N appears to be a reliable indicator of trophic level withinrocky shore ecosystems and a biogeographic enrichment innitrogen indicates strong ties to regional oceanography andnutrient availability. However, d13C analysis is more problematic.While the importance of macroalgal-derived organic carbon infilter feeder diets is clear, the links among living macroalgae,nearshore SPM and consumers are still unclear. Problems arisebecause of variation in macroalgal signatures, which are influencedby several factors, omnivory, and the way in which algal productionenters the food web.

Stable isotope analysis is highly effective in the analysis ofpelagic ecosystems, in which primary production is normallygrazed directly. It is more difficult to apply to benthic systems.Much of this difficulty stems from primary production largelyentering the benthic food web indirectly, as detritus, in contrast topelagic ecosystems. Given this, the clear biogeographic trendsshown by filter feeders are striking and demonstrate very strongsignatures of the physical environment.

Acknowledgements

Many thanks to J. Lanham for isotopic analysis at the Stable LightIsotope Unit, University of Cape Town, South Africa. This study wassupported by funds from the National Research Foundation (NRF)and the African Coelacanth Ecosystem Programme (ACEP). All ofthis help is gratefully acknowledged.

J.M. Hill, C.D. McQuaid / Estuarine, Coastal and Shelf Science 80 (2008) 261–268268

References

Abend, A., Smith, T., 1997. Differences in stable isotope ratios of carbon and nitrogenbetween long-finned pilot whales (Globicephala melas) and their primary preyin the western north Atlantic. ICES Journal of Marine Science 54, 500–503.

Blanchette, C., Broitman, B., Gaines, S., 2006. Intertidal community structure andoceanographic patterns around Santa Cruz Island, CA, USA. Marine Biology 149,689–701.

Bode, A., Carrera, P., Lens, S., 2003. The pelagic foodweb in the upwelling ecosystemof Galicia (NW Spain) during spring: natural abundance of stable carbon andnitrogen isotopes. ICES Journal of Marine Science 60, 11–22.

Branch, G., 1971. The ecology of Patella Linnaeus from the Cape Peninsula, SouthAfrica I. Zonation, movements and feeding. Zoologica Africa 6, 1–38.

Burkhardt, S., Riebesell, U., Zondervan, I., 1999a. Effects of growth rate, CO2concentration, and cell size on the stable carbon isotope fractionation in marinephytoplankton. Geochimica et Cosmochimica Acta 63, 3729–3741.

Burkhardt, S., Riebesell, U., Zondervan, I., 1999b. Stable carbon isotope fractionationby marine phytoplankton in response to daylength, growth rate, and CO2availability. Marine Ecology Progress Series 184, 31–41.

Bustamante, R., Branch, G., 1996. The dependence of intertidal consumers on kelp-derived organic matter on the west coast of South Africa. Journal of Experi-mental Marine Biology and Ecology 196, 1–28.

Bustamante, R.H., Branch, G.M., Eekhout, S., Robertson, B., Zoutendyk, P.,Schleyer, M., Dye, A., Keats, D., Jurd, M., McQuaid, C.D., 1995. Gradients ofintertidal productivity around the coast of South Africa and their relationshipwith consumer biomass. Oecologia 102, 189–201.

Carpenter, S.R., Kitchell, J.F., Hodgson, J.R., 1985. Cascading trophic interactions andlake productivity. Bioscience 35, 634–639.

Carpenter, S.R., Cole, J.J., Hodgson, J.R., Kitchell, J.F., Pace, M.L., Bade, D.,Cottingham, K.L., Essington, T.E., Houser, J.N., Schindler, D.E., 2001. Trophiccascades, nutrients, and lake productivity: whole-lake experiments. EcologicalMonographs 71, 163–186.

Chen, B., Wise, D.H., 1999. Bottom-up limitation of predaceous arthropods ina detritus-based terrestrial food web. Ecology 80, 761–772.

Chu, F.-L., Soudant, P., Cruz-RodroAguez, L., Hale, R., 2000. PCB uptake and accu-mulation by oysters (Crassostrea virginica) exposed via a contaminated algaldiet. Marine Environmental Research 50, 217–221.

De Clerck, O., Bolton, J., Anderson, R., Coppejans, E., 2005. Guide to the Seaweeds ofKwaZulu-Natal. National Botanic Garden of Belgium, Meise, Belgium, 294 pp.

DeNiro, M., Epstein, S., 1978. Influence of diet on the distribution of carbon isotopesin animals. Geochimica et Cosmochimica Acta 42, 495–506.

Descolas-Gros, C., Fontugne, M., 1985. Carbon fixation in marine phytoplankton.Carboxylase activities and stable carbon-isotope ratios; physiological andpaleoclimatological aspects. Marine Biology 87, 1–6.

Dunton, K., 2001. d15N and d13C measurements of Antarctic Peninsula fauna: trophicrelationships and assimilation of benthic seaweeds. American Zoologist 41,99–112.

Fredriksen, S., 2003. Food web studies in a Norwegian kelp forest based on stableisotope (d13C and d15N) analysis. Marine Ecology Progress Series 260, 71–81.

Fretwell, S.D., 1977. The regulation of plant communities by food chains exploitingthem. Perspectives in Biology and Medicine 20, 169–185.

Fry, B., Sherr, E., 1984. d13C measurements as indicators of carbon flow in marineand freshwater ecosystems. Contributions in Marine Science 27, 13–17.

Gorokhova, E., Hansson, S.,1999. An experimental study on variations in stable carbonand nitrogen isotope fractionation during growth of Mysis mixta and Neomysisinteger. Canadian Journal of Fisheries and Aquatic Sciences 56, 2203–2210.

Hansson, S., Hobbie, J., Elmgren, R., Larsson, U., Fry, B., Johansson, S., 1997. The stablenitrogen isotope ratio as a marker of food-web interactions and fish migration.Ecology 78, 2249–2257.

Herendeen, R.A., 2004. Bottom-up and top-down effects in food chains depend onfunctional dependence: an explicit framework. Ecological Modelling 171, 21–33.

Hill, J., McQuaid, C., Kaehler, S., 2006. Biogeographic and nearshore/offshore trendsin isotope ratios of intertidal mussels and their food sources around the coast ofsouthern Africa. Marine Ecology Progress Series 318, 63–73.

Lasiak, T., 2006. Spatial variation in density and biomass of patellid limpets insideand outside a marine protected area. Journal of Molluscan Studies 72, 137–142.

Laws, E., Bidigare, R., Popp, B., 1997. Effect of growth rate and CO2 concentration oncarbon fractionation by the marine diatom Phaeodactylum tricornutum.Limnology and Oceanography 42, 1552–1560.

Marquis, R.J., Whelan, C.J., 1994. Insectivorous birds increase growth of white oakthrough consumption of leaf-chewing insects. Ecology 75, 2007–2014.

McQuaid, C., 1985. Seasonal variation in biomass and zonation of nine intertidalalgae in relation to changes in radiation, sea temperature and tidal regime.Botanica Marina XXVIII, 539–544.

McQuaid, C., Froneman, P., 1993. Mutualism between the territorial intertidal limpetPatella longicosta and the crustose alga Ralfsia verrucosa. Oecologia 96, 128–133.

McQuaid, C., Lindsay, T., 2000. Effect of wave exposure on growth and mortalityrates of the mussel Perna perna: bottom-up regulation of intertidal populations.Marine Ecology Progress Series 206, 147–154.

McQuaid, C., Payne, A., 1998. Regionalism in marine biology: the convergence ofecology, economics and politics in South Africa. South African Journal of Science94, 433–436.

Menge, B., 1992. Community regulation: under what conditions are bottom-upfactors important on rocky shores? Ecology 73, 755–765.

Menge, B., 2000. Top-down and bottom-up community regulation in marine rockyintertidal habitats. Journal of Experimental Marine Biology and Ecology 250,257–289.

Menge, B., Daley, B., Lubchenco, J., Sanford, E., Dahlhoff, E., Halpin, P., Hudson, G.,Burnaford, J., 1999. Top-down and bottom-up regulation of New Zealand rockyintertidal communities. Ecological Monographs 69, 297–330.

Menge, B., Farrell, T., 1989. Community structure and interaction webs in shallowmarine hard-bottom communities tests of an environmental stress model.Advances in Ecological Research 19, 189–262.

Menge, B., Lubchenco, J., Bracken, M., Chan, F., Foley, M., Freidenburg, T., Gaines, S.,Hudson, G., Krenz, C., Leslie, H., Menge, D., Russell, R., Webster, M., 2003. Coastaloceanography sets the pace of rocky intertidal community dynamics.Proceedings of the National Academy of Sciences of the United States ofAmerica 100, 12229–12234.

Minagawa, M., Wada, E., 1984. Stepwise enrichment of 15N along food chains:further evidence and the relation between 15N and animal age. Geochimica etCosmochimica Acta 8, 1135–1140.

Miyake, Y., Wada, E., 1967. The abundance ratio of 15N/14N in marine environments.Records of Oceanographic Works in Japan 9, 32–53.

Monterio, P., James, A., Sholto-Douglas, A., Field, J., 1991. The d13C trophic positionisotope spectrum as a tool to define and quantify carbon pathways in marinefood webs. Marine Ecology Progress Series 78, 33–40.

Moore, S., Suthers, I., 2005. Can the nitrogen and carbon stable isotopes of thepygmy mussel, Xenostrobus securis, indicate catchment disturbance for estu-aries in Northern New South Wales, Australia? Estuaries 28, 714–725.

Nielsen, K., Navarrete, S., 2004. Mesoscale regulation comes from the bottom-up:intertidal interactions between consumersand upwelling. Ecology Letters 7, 31–41.

O’Reilly, C., Hecky, R., Cohen, A., Plisnier, P.-D., 2002. Interpreting stable isotopes infood webs: recognizing the role of time averaging at different trophic levels.Limnology and Oceanography 47, 306–309.

Page, H.M., Lastra, M., 2003. Diet of intertidal bivalves in the R�ya de Arosa (NWSpain): evidence from stable C and N isotope analysis. Marine Biology 143,519–532.

Peterson, B., Howarth, R., 1987. Sulfur, carbon, and nitrogen isotopes used to traceorganic matter flow in the salt-marsh estuaries of Sapelo Island, Georgia.Limnology and Oceanography 32, 1195–1213.

Phillips, D.L., Gregg, J.W., 2003. Source partitioning using stable isotopes: copingwith too many sources. Oecologia 136, 261–269.

Post, D., 2002. Using stable isotopes to estimate trophic position: models, methodsand assumptions. Ecology 83, 703–718.

Power, M., 1992. Top-down and bottom-up forces in food webs:do plants haveprimacy? Ecology 73, 733–746.

Raikow, D., Hamilton, S.K., 2001. Bivalve diets in a midwestern U.S. stream: a stableisotope enrichment study. Limnology and Oceanography 46, 514–522.

Saino, T., Hattori, A., 1980. 15N natural abundance in oceanic suspended particulatematter. Nature 283, 752–754.

Sara, G., Scilipoti, D., Mazzola, A., Modica, A., 2004. Effects of fish farming waste tosedimentary and particulate organic matter in a southern Mediterranean area(Gulf of Castellammare, Sicily): a multiple stable isotope study (d13C and d15N).Aquaculture 234, 199–213.

Seitz, R., Lipcius, R., 2001. Variation in top-down and bottom-up control ofmarine bivalves at differing spatial scales. ICES Journal of Marine Science 58,689–699.

Sharkey, T., Berry, J., 1985. Carbon isotope fractionation of algae as influenced by aninducible CO2 concentrating mechanisms. In: Lucas, W.J., Berry, J.A. (Eds.),Inorganic Carbon Uptake by Aquatic Photosynthetic Organisms. AmericanSociety of Plant Physiologists, Rockwood, MD, pp. 389–401.

Thompson, P., Calvert, S., 1994. Carbon isotope fractionation by a marine diatom:the influence of irradiance, daylength, pH, and nitrogen source. Limnology andOceanography 39, 1835–1844.

Tieszen, L., Boutton, T.W., Tesdahl, K., Slade, N., 1983. Fractionation and turnover ofstable carbon isotopes in animal tissues: implications for d13C analysis of diet.Oecologia 57, 32–37.

Wieters, E.A., 2005. Upwelling control of positive interactions over mesoscales:a new link between bottom-up and top-down processes on rocky shores.Marine Ecology Progress Series 301, 43–54.

Wong, W., Sackett, W., 1978. Fractionation of stable carbon isotopes by marinephytoplankton. Geochimica et Cosmochimica Acta 42, 1809–1815.