Changes in trophic level of Squatina guggenheim with increasing body length: relationships with type, size and trophic level of its prey

Changes in trophic level of Squatina guggenheimwith increasing body length: relationships with type, sizeand trophic level of its prey

Rodolfo Vögler & Andrés C. Milessi &Luis O. Duarte

Received: 9 April 2007 /Accepted: 24 July 2008 / Published online: 27 August 2008# Springer Science + Business Media B.V. 2008

Abstract The occurrence of changes in the trophiclevel (TL) of sharks with growth has not beenquantified until now. Here length-related changes onSquatina guggenheim Marini trophic level weredetermined, and shifts in type, size and trophic levelof its prey were analysed. Sampling took place duringfive bottom trawl surveys conducted in the Argentine–Uruguayan Common Fishing Zone during spring(December/1995, October/1997) and fall (March/1997, March–April/1998, May–June/1998), using anEngel bottom-trawl net to capture the sharks. Threelength groups were defined based on diet compositionand using a cluster analysis (group I, 23–60 cm; groupII, 61–80 cm; group III, 81–91 cm LT). An ANOSIMprocedure detected significant differences (P<0.05) inthe diet spectrum between the three length groups. The

smallest sharks (group I) ingested fish prey rangingfrom 5 to 21 cm LT, medium sharks (group II) fed onfish prey between 11 and 35 cm LT, and largest sharks(group III) preyed on fish between 13 and 40 cm LT.Diet structure of length groups were discriminated byalmost the same prey taxa that characterized them. Theincrease of S. guggenheim body length promoted adecrease in the relative importance of small pelagicfishes. Contrarily, prey as medium benthopelagicfishes, medium pelagic squid and medium benthope-lagic fishes showed an inverse tendency, indicating abroad diet spectrum of adults. Predator-length andprey-length relationship indicated a trend where 44.8%of S. guggenheim diet was integrated by prey <20% oftheir own body length and 32.8% of their diet wascomposed by prey >30% of their own length. Theincrease of mean prey weight was associated withthe increase of predator weight and length. Smallestsharks (group I) were identified as secondary consum-ers (TL<4) whereas medium sharks (group II) andlargest sharks (group III) were placed as tertiaryconsumers (TL>4). The study revealed an increase inS. guggenheim TL with shark growth as a consequenceof changes on type, size and TL of prey ingested.

Keywords Angular angel shark . Feeding ecology.

Predator–prey interactions . Uruguay.Marine foodwebs

Introduction

Sharks are typical top predators in marine ecosystems(Stevens et al. 2000), playing a major role in the

Environ Biol Fish (2009) 84:41–52DOI 10.1007/s10641-008-9387-x

R. Vögler (*)Programa de Doctorado,Centro Interdisciplinario de Ciencias Marinas,Av. Instituto Politécnico Nacionals/n. Col. Playa Palo de Santa Rita,CP 952 La Paz, Baja California Sur, Méxicoe-mail: [email protected]

A. C. MilessiConsejo Nacional de InvestigacionesCientíficas y Técnicas (CONICET),Rivadavia 1917,Buenos Aires, Argentina

L. O. DuarteLaboratorio de Investigaciones Pesqueras Tropicales,Universidad del Magdalena,Cra. 32 # 22-08,Santa Marta, Colombia

energy fluxes exchange between upper and lowertrophic levels (Wetherbee and Cortés 2004). In spiteof growing interest in shark feeding ecology over thelast decade (e.g. Brewer et al. 1995; Cortés et al.1996; Ellis et al. 1996; White et al. 2004), the firststudy that estimated the position of 149 shark speciesin marine food webs (Cortés 1999) was accomplishedonly recently. Numerous studies have found that bothtype and size of prey ingested change with increasingpredator size (Lyle 1983; Werner and Gilliam 1984;Platell et al. 1998; Schefer et al. 2002; Cherel andDuhamel 2004; Wetherbee and Cortés 2004). How-ever, diet shifts are more often reported qualitativelyrather than based on rigorous statistical analysis.

Squatina guggenheim is one of the most wide-spread endemic elasmobranchs inhabiting bottomwaters of the Southwest Atlantic (33°–41° S) coastalecosystem (Vögler et al. 2008). This bottom-dwellingshark is an important fishing resource caught off thecoastal zones of Uruguay (Paesch and Meneses1999), southern Brazil (Villwock and Vooren 2003)and northern Argentina (Chiaramonte 1998). Theinformation about the diet and feeding habits of S.guggenheim along their geographic range remainsfragmentary. Vögler et al. (2003) have studied someaspects of S. guggenheim trophic ecology, such as dietdiversity and composition, feeding strategy (by sexand size), and diet variability (spatial and seasonal)within the shelf of the Argentine–Uruguayan Com-mon Fishing Zone (AUCFZ). However, changes on S.guggenheim trophic level with growth as well aschanges on the trophic level of their prey remainunknown. In addition, the relationship of thesechanges with type, size and trophic level of preyingested have not been determined until now.

Accordingly, the aims of the present study were:(1) to determine changes in the trophic level of S.guggenheim with increasing body length and (2) toanalyze shifts on the type, size and trophic level ofprey ingested during the growth of S. guggenheim.

Materials and methods

Sample collection

Individuals of S. guggenheim were collected from fivebottom trawl surveys conducted in spring (December1995, October 1997) and fall (March 1997, March–

April 1998, May–June 1998) by the DirecciónNacional de Recursos Acuáticos (DINARA) onboardthe RV “Aldebarán”. The surveys used a stratifiedrandom sample design and covered most of theAUCFZ (34°00′–39°30′ S; 51°10′–59°10′ W) with abathymetric range between 3.5 and 266.0 m (Fig. 1).Standard bottom trawls were performed in all surveyswith an Engel trawl net (100 mm mesh in the wingsand 60 mm mesh in the cod ends, 4 m verticalopening and 15 m horizontal aperture). Each fishingtow lasted 30 min (μ=30, σ=2.40) at a speed of 3.0knots (μ=2.98, σ=0.14). Only daytime sampling wasconducted (from 05:30 to 19:30 h).

A total of 1,280 S. guggenheim specimens werecaptured, counted and weighed (±0.1 kg) on board.During all surveys, 947 S. guggenheim individualswere measured to the lowest cm of total body length(LT), their sex was recorded and the contents of theirstomachs were analyzed. The prey were separated andidentified to the lowest possible taxonomic levelusing standard taxonomic keys (Menni et al. 1984;Boschi et al. 1992). Only bony and cartilaginous fishprey of S. guggenheim were measured to the lowestcm of LT.

Numerical analysis

Cluster analysis was performed to divide the total S.guggenheim sample in length intervals of 5 cm eachusing the unweighted pairwise group mean averagemethod (UPGMA) with the Euclidean distance as ameasure of dissimilarity. Euclidean distance betweensize groups was calculated as:

D X ;Yð Þ ¼Xp

i¼1Xj � Yj� �2� �

1=2 ð1Þ

where D(X,Y) was the distance between two lengthgroups X and Y; p was the number of length groups;Xj and Yj were the relative importance of prey item j atthe length groups X and Y respectively.

A Monte Carlo-based ANOSIM procedure (Clarkeand Green 1988) was used to test the null hypothesisof no differences in the diet composition between S.guggenheim length groups. This test is based on therank similarities between samples in the underlyingsimilarity matrix (Clarke and Green 1988). Thetypifying prey and discriminating prey for each lengthgroup were determined using the similarity percentprocedure (SIMPER, Clarke 1993). This procedure

42 Environ Biol Fish (2009) 84:41–52

determines the average contribution of each prey tothe similarity (typifying species) and dissimilarity(discriminating species) between length groups. Bonyfish remains and unidentified digested remains wereexcluded from these analyses.

Squatina guggenheim prey were grouped in func-tional categories according to taxonomic group(mollusc, annelid, crustacean, fish), size (small,medium, large) and habit (pelagic, benthopelagic,demersal) (Table 1). This procedure included eachone of the 32 prey identified by Vögler et al. (2003).

The contribution of each prey to the diet of S.guggenheim was determined by two Relative Meas-ures of Prey Quantity (RMPQ, after Assis 1996):numerical index (Ni) and frequency of occurrenceindex (Fi) (Hyslop 1980). A generalized form of theRelative Importance Index (RI) (George and Hadley1979) was computed for each prey taxon as:

RI ¼ 100Xm

i¼1Vij

.Xm

i¼1

Xn

j¼1Vij

� �ð2Þ

where Vij is the i-th RMPQ of prey j, n is the numberof individuals included in each prey taxon, and m isthe number of different prey taxa.

The RI of the functional prey categories wasplotted for each length group of S. guggenheim. Thelevel of importance of each prey category wasestablished according with discontinuities in the slopeof the RI curve. The categories of prey with high RIvalues were assigned at the first level of importanceand so forth.

The trophic level (TL) was calculated for eachlength group of S. guggenheim applying the method-ology proposed by Cortés (1999) as:

TL ¼ 1þXn

j¼1Pj � TLj

� �ð3Þ

where TL is the trophic level of S. guggenheim(predator), Pj is the proportion of prey taxon j in thepredator stomach, TLj is the trophic level of each preytaxon j and n is the number of prey taxon recorded inthe predator stomach.

In the present study, Pj correspond to numericalindex (Ni) values of each prey estimated by Vögler etal. (2003) from S. guggenheim stomach contentanalysis. TLj values for each prey were obtained fromthe Sea Around Us Project (Sea Around Us 2006),FishBase (Froese and Pauly 2006) and CephBase(Wood and Day 2006) as of December 2006 (Table 1).

Fig. 1 a The location of theArgentine–UruguayanCommon Fishing Zone(AUCFZ) in South Amer-ica. b Study area situatedinside the AUCFZ. Isobathsof 50, 100, 200 and 400 mare shown. Dots indicatelocation of fishing tows

Environ Biol Fish (2009) 84:41–52 43

According to Vögler et al. (2003) cannibalism in S.guggenheim occurred from larger males to smallestindividuals (of both sexes). Then, the TL of S.guggenheim, as a prey, was estimated for the smallest

length group (group I, 23–60 cm) using the method-ology proposed by Cortés (1999).

The TL for the whole population of S. guggenheimwas computed as the average TL estimated for each

Table 1 Trophic level, common name and functional category of Squatina guggenheim prey, listed in descending order of trophic level

Prey items Common name TL FC

MolluscaIllex argentinus (Castellanos) Argentine shortfin squid 3.20a MPSLoligo sanpaulensis Brakoniecki Sao Paulo squid 3.20d SPSOctopus tehuelchus Orbigny Tehuelche octopus 3.20d SPSBivalvia (unidentified) Benthic mollusc 2.10b SBMGastropoda (unidentified) Benthic mollusc 2.10b SBMAnnelidaPolychaeta (unidentified) Bristle worms 2.50b SBAAphrodita spp. Linnaeus Sea mouse 2.50b SBACrustaceaPeltarion spinosulum (White) Tractor crab 2.52b SBCPleoticus muelleri (Bate) Argentine red shrimp 2.52b SDCPenaeidae (unidentified) Shrimp 2.52b SDCActinopterygiiPercophis brasiliensis Quoy and Gaimard Brazilian flathead 4.40a MDFParalichthys spp. Girard Flounder 4.35b MDFGenypterus brasiliensis Regan Cusk-eels 4.34a LDFGenypterus blacodes (Forster) Cusk-eels 4.34a LBDFCottoperca gobio (Günther) Perch-likes 4.26a LDFPrionotus nudigula Ginsburg Red searobin 4.20a SDFMerluccius hubbsi Marini Argentine hake 4.08a LBPFCynoscion guatucupa (Cuvier) Striped weakfish 3.90b MBPFHelicolenus dactylopterus lahillei (Norman) Blackbelly rosefish 3.81a MDFUrophycis brasiliensis (Kaup) Brazilian codling 3.79a MDFPrionotus punctatus (Bloch) Bluewing searobin 3.77a MDFBatrachoididae Toadfishes 3.67a MDFDules auriga (Cuvier) 3.60a SBPFRaneya spp. (Kaup) Cusk-eels 3.56b MDFPatagonotothen longipes (Steindachner) Cod icefish 3.49a MBPFPatagonotothen ramsayi (Regan) Cod icefish 3.49a MBPFTrachurus lathami Nichols Rough scad 3.45a MPFConger orbignyanus Valenciennes Argentine conger 3.40a LDFParalonchurus brasiliensis (Steindachner) Banded croaker 3.06a SDFUmbrina canosai Berg Argentine croaker 2.80a MDFEngraulis anchoita Hubbs and Marini Argentine anchovy 2.48a SPFChondrichthyesSquatina guggenheim Marini [23–60 cm LT] Squatina guggenheim 3.69c LDF

TL trophic level, FC functional category, SBA small benthic annelids, SBC small benthic crustacean, SBM small benthic mollusc,SBPF small benthopelagic fish, SDC small demersal crustacean, SDC small demersal fish, SPF small pelagic fish, SPS small pelagicsquid, MBPF medium benthopelagic fish, MDF medium demersal fish, MPF medium pelagic fish, MPS medium pelagic squid, LBDFlarge bathydemersal fish, LBPF large benthopelagic fish, LDF large demersal fisha Froese and Pauly (2006)b Sea Around Us (2006)c Current studydWood and Day (2006)

44 Environ Biol Fish (2009) 84:41–52

length group and weighted by their relative proportionat the length frequency distribution during thesampling period, as:

TLp ¼X

TLi fi ð4Þwhere TLp is the trophic level of the population, TLi

is the trophic level of the i-th length group and fi is therelative proportion of the i-th length group within thetotal specimens sampled during the study period.

A scatter diagram was plotted to evaluate therelationship between the lengths of prey and predatorat the entire body length spectrum of S. guggenheim.The Spearman test (Zar 1999) was performed toestablish the significance level (P<0.01) of theregression between the lengths of prey and predator.According with Scharf et al. (2000), prey length datawere converted to a ratio scale by dividing each preylength by its corresponding predator length and thenplotting these prey length/predator length ratios. Therelative frequency distribution and cumulative fre-quency distribution of prey length/predator lengthratios were constructed to evaluate the range of fishprey length eaten by S. guggenheim at their entirebody length spectrum. Finally, the total body length(LT) was converted to total body weight (WT) of prey

using length/weight relationships specific to each fishprey as:

WT ¼ a� LbT ð5Þwhere a is the factor and b is the exponent.

The values of the parameters a and b whereobtained from FishBase home page (Froese and Pauly2006) for each fish prey. The same procedure wasapplied to S. guggenheim as predator, however thevalues of a and b were obtained from Colonello et al.(2007). Then, the weights of fish prey were plottedagainst predator weight. Only bony fish and cartilag-inous fish prey of S. guggenheim were utilised toestablish these analyses (Table 2). It was unable toestablish the minimum and maximum weights ofParalichthys spp. and Raneya spp., because it was notpossible to make a distinction between species withinthose genera. In turn, data for length and weight ofHelicolenus dactylopterus lahillei (Norman), Dulesauriga (Cuvier) and Cottoperca gobio (Günther) werenot included because the LT of these prey was notmeasured during sampling on board. The Spearmantest (Zar 1999) was performed to establish thesignificance level (P<0.01) of the regression betweenthe weights of prey and predator.

Table 2 Marine fish prey examined in this study from the stomach contents of Squatina guggenheim

Fish prey LT min (cm) LT max (cm) WT min (g) WT max (g)

Engraulis anchoita 7 14 2.81 31.6Squatina guggenheim 29 29 184.62 –Urophycis brasiliensis 35 35 354.95 –Conger orbignyanus 34 34 36.71 –Paralonchurus brasiliensis 10 20 6.61 81.65Merluccius hubbsi 19 31 68.33 231.92Macrodon ancylodon 7 34 1.88 329.98Cynoscion guatucupa 5 25 3.13 –Percophis brasiliensis 28 40 90.19 262.68Trachurus lathami 8 11 4.68 11.9Prionotus punctatus 12 12 22.8 –Prionotus nudigula 19 19 78.29 –Genypterus spp. 20 25 19.62 43.87Paralichthys spp. 8 15 – –Patagonotothen spp. 16 20 – –Raneya spp. 17 17 – –Cottoperca gobio – – – –Dules auriga – – – –Helicolenus dactylopterus lahillei – – – –

LT min minimum total body length, LT max maximum total body length, WT min minimum total body weight, WT max maximum totalbody weight

Environ Biol Fish (2009) 84:41–52 45

Results

A total of 947 stomachs of S. guggenheim wereexamined during the study period at sea and 58.7%(n=552) of them contained food.

Three length groups were defined by the clusteranalysis, they were based on diet composition: groupI (23–60 cm LT), group II (61–80 cm LT) and groupIII (81–91 cm LT) (Fig. 2). Significant differences(P<0.05) in the diet spectrum were detected by theANOSIM procedure when compared group I vs II(R=0.87, P=0.005) and group I vs III (R=0.99, P=0.036). No significant differences (P>0.05) weredetected between the diet of group II and III (R=0.96, P=0.67).

Grouping the prey in functional categories showedthat small pelagic fishes (i.e. Engraulis anchoitaHubbs and Marini) were the predominant prey itemin the diet of smallest individuals of S. guggenheim(group I) and their relative importance decreased inthe diet as the body length of predator increase(Fig. 3). In contrast, medium benthopelagic fisheswere the most important prey of larger sharks (groupsII and III) (Fig. 3). Small demersal crustaceans(shrimps and crabs) had a high relative importancein the diet of sharks included in group II. Mediumpelagic squid (i.e. Illex argentinus (Castellanos)) andmedium demersal fishes (e.g. Helicolenus dactylopte-rus lahillei (Norman)) represented secondarily impor-tant prey groups in the diet of the largest sharks(group III) (Fig. 3).

SIMPER analysis revealed that the diet of thesmallest individuals of S. guggenheim (group I) were

typified by E. anchoita and Patagonotothen longipes(Steindachner), while in the case of medium sharks(group II) the typified prey were shrimps (Penaei-dae), Cynoscion guatucupa (Cuvier) and Merlucciushubbsi Marini. The diet of the largest sharks (groupIII) was characterised by Patagonotothen ramsayi(Regan) and I. argentinus (Table 3). The dietstructure of length groups of S. guggenheim wasdiscriminated by almost the same prey taxa thatcharacterized them. Thus, E. anchoita consistentlydiscriminated the diet between the smallest sharks(group I) and other length groups. In contrast,shrimps (Penaeidae) were the main discriminatingprey between the medium sharks (group II) and theother length groups. Finally, I. argentinus and P.ramsayi represented the main discriminating preybetween the largest sharks (group III) and the otherlength groups (Table 4).

Fig. 2 Squatina guggenheim length groups based on dietcomposition and obtained by a cluster analysis. I, 23–60 cm; II,61–80 cm; III, 81–91 cm LT

Fig. 3 Relative importance (RI) of functional prey categories(FC) in the diet of Squatina guggenheim length groups (groupI, 23–60 cm; group II, 61–80 cm; group III, 81–91 cm LT).Dashed vertical lines indicate the three levels of importance ofthe functional prey categories according with discontinuities inthe slope of the RI curve. The categories of prey with high RIvalues were assigned at the first level of importance and soforth. Note the different scales in the y-axis between lengthgroups, within the minimum (0) and maximum (100) values ofRI. FC abbreviations as showed in Table 1

46 Environ Biol Fish (2009) 84:41–52

The size range of fish prey eaten by S. guggen-heim expanded with increasing predator body length.Smallest sharks (group I) ingested fish prey rangingfrom 5 to 21 cm LT (mean=11 cm LT), mediumsharks (group II) fed on fish prey between 11 and35 cm LT (mean=21 cm LT), while largest sharks(group III) consumed fish prey between 13 and40 cm LT (mean=24 cm LT). The positive relation-ship between the length of fish prey and the length ofS. guggenheim was significant (rs=0.67, P<0.01)(Fig. 4a). The asymmetric relative frequency distri-bution of prey length/predator length ratios wasreflected by two trends where 44.8% and 32.8% ofS. guggenheim diet was composed by prey of <20%and >30% of their own body length, respectively(Fig. 4b). Finally, the positive relationship betweenthe weight of fish prey and the weight of S.guggenheim was significant (rs=0.57, P<0.01)(Fig. 4c).

The trophic level of S. guggenheim increased withincreasing body length. Sharks of group I wereidentified as secondary consumers (TL<4) whereasindividuals of groups II and III were tertiary consum-ers (TL>4) (Table 5). The estimated TL for the wholepopulation of S. guggenheim was 3.90.

Discussion

The present study revealed that the relative impor-tance of small pelagic fishes (i.e. E. anchoita)decreased in the diet of S. guggenheim as predatorbody length increases, while medium benthopelagicfishes (C. guatucupa, P. ramsayi) showed an inverse

tendency. Furthermore, medium pelagic squid (I.argentinus) and medium demersal fishes were impor-tant prey categories in the diet of the largestindividuals (group III), indicating a broad preyspectrum of S. guggenheim adults. Also, changes inthe TL of prey ingested were detected and they wererelated with functional characteristics (type, size, lifehabits) of each prey. According to Vögler et al. (2003)bony fishes were the main prey in the diet of the threeshark length groups, however, an increase in dietdiversity was observed during the growth of S.guggenheim. In many shark species, the increase ofbody length promotes multiple changes related withmovement’s patterns, swimming speed, size of jaws,teeth and stomachs, energy requirements, experiencein prey handling, as well as changes in the habitatselection and vulnerability to predation (Wetherbeeand Cortés 2004). Increase in body length hadpromoted changes in the type and size of preyingested by S. guggenheim and also changes ofpredator feeding habits were observed.

Vögler et al. (2003) observed spatial and temporalvariability in S. guggenheim diet diversity. The high-est diet diversity was detected in the sharks fromshallow depths during spring. In contrast, duringautumn the highest diet diversity was observed insharks from greater depths (Vögler et al. 2003).Therefore, the evidence suggests that prey availabilitydiffers in space (between depths) and time (betweenseasons) within the shelf of the AUCFZ. In thiscontext, E. anchoita was the main prey of S.guggenheim, as well as I. argentinus and M. hubbsi

Table 3 Percentage of similarity of typifying (over 5%) preytaxon for the three length groups of Squatina guggenheim

Typifying prey taxon I II III

Engraulis anchoita 43.26a 16.50 12.97Patagonotothen longipes 16.40a

Penaeidae 5.68 23.29a

Cynoscion guatucupa 5.56 22.25a 14.65Merluccius hubbsi 17.29a 14.65Patagonotothen ramsayi 14.15 21.42a

Illex argentinus 16.47a

a Prey taxon with higher contribution to the similarity (mean/SD>4) percentage. I, 23–60 cm; II, 61–80 cm; III, 81–91 cm LT

Table 4 Percentage of dissimilarity of discriminating (over5%) prey taxon for the three length groups of Squatinaguggenheim

Discriminating prey taxon I vs II I vs III II vs III

Engraulis anchoita 11.81a 6.31a

Patagonotothen longipes 7.14a 8.79a

Penaeidae 6.96a 7.25a

Cynoscion guatucupa 7.25a 5.45Merluccius hubbsi 6.42a 5.42Bivalvia 6.49 5.48Patagonotothen ramsayi 7.99 6.46a 12.37a

Illex argentinus 8.70a 6.86a

Helicolenus dactylopterus lahillei 7.79a 6.78a

a Prey taxon with higher contribution to the dissimilarity (mean/SD>2) percentage. I, 23–60 cm; II, 61–80 cm; III, 81–91 cm LT

Environ Biol Fish (2009) 84:41–52 47

were secondary preys (Vögler et al. 2003). Theseasonal distribution of these species within theAUCFZ was related to their reproductive biology.For example, E. anchoita spawns from September to

November with some spawning activity in thesouthern coastal zone of the continental shelf atdepths <100 m in the Buenos Aires (Argentina)sector of the AUCFZ (Bakun and Parrish 1991).Merluccius hubbsi spawns between autumn andwinter in the north of the AUCFZ (Norbis et al.,1999), and I. argentinus spawns in this same areaduring winter and spring (Bazzino and Quiñones1999). The AUCFZ is influenced by four watermasses (Acha et al. 2004), namely: the Brazil Current(warm water); the Malvinas Current (cold water); thePatagonia Current (cold water), and Río de la Platadischarges (freshwater) (Guerrero et al. 1997). TheBrazil (Subtropical) and Malvinas (SubAntarctic)Currents meet, on average, at 36° S, generating the‘Subtropical Convergence’, a large permanent frontalsystem of high temporal variability that is displacednorthward in winter and southward in summer (Olsonet al. 1988). The flow of the Río de la Plata changesseasonally discharging freshwater along the Uru-guayan coast during winter and along the Argentineancoast in spring and summer (Guerrero et al. 1997).The dynamic interaction of these water massespromotes seasonal and spatial variability of tempera-ture and salinity in the AUCFZ, which couldinfluence the distribution of S. guggenheim and theirprey, promoting an overlap between them both inspace and in time.

Vögler et al. (2003) operationally grouped S.guggenheim specimens in three length groups. Here,also three length-related groups were elucidated usinga formal statistical procedure. The length groupsdefined in the two studies were different, but bothhave detected changes of prey relative importancewith the increase of S. guggenheim body length.

Fig. 4 a Lineal regression between total body length of prey(prey LT, cm) and total body length of predator (predator LT,cm) at the entire length spectrum of Squatina guggenheim. bRelative frequency distribution (bars) and cumulative frequen-cy distribution (continuous line, filled dots at 5% intervals) ofprey length/predator length ratios. c Lineal regression betweentotal body weight of prey (prey WT, g) and total body weight ofpredator (predator WT, g) at the entire weight spectrum of S.guggenheim. At a and c dashed lines indicate ±95% confidenceinterval of Spearman test (rs)

Table 5 Trophic level of the three length groups of Squatinaguggenheim

Length group TL N n

I 3.69 342 415II 4.26 145 199III 4.40 65 93

I, 23–60 cm; II, 61–80 cm; III, 81–91 cm LTTL Trophic level N number of stomachs with food contents, ntotal number of prey found in the stomachs of each lengthgroup

48 Environ Biol Fish (2009) 84:41–52

The feeding strategy that characterizes manycarnivorous fishes reflects that prey size increase withthe increase of predator body length (Karpouzi andStergiou 2003). This kind of relationship is consistentwith an asymptotic relationship between TL and LT offish (Cortés 1999; Stergiou and Karpouzi 2002).However, in the case of S. guggenheim the asymmet-ric relative frequency distribution of prey length/predator length ratios suggest that this predatorincorporates a considerable number of intermediatelength prey in their diet while it also continues to feedon several small prey during the growth. Scharf et al.(2000) have demonstrated that this kind of preylength/predator length relationship is a commoncharacteristic of many fish predators’ diet andcontrasts with optimal foraging theory (Stephens andKrebs 1986), which predicts that the largest preyavailable should be eaten to maximize feedingefficiency.

It is important to note that the present analysis onlyincluded teleost and cartilaginous fish prey. Through-out the growth, the diet of S. guggenheim was ingeneral more concentrated on fish, rather than oninvertebrates (Vögler et al. 2003) but significantdifferences (P≤0.05) between the diet of three lengthgroups were found. Further studies are necessary toelucidate the cause of the differences in the dietbetween length groups with the incorporation oflength and weight of other prey groups (i.e. crusta-ceans, molluscs and annelids).

Food and feeding habits determine the position ofanimals within food webs, and define their ecologicalrole (Pauly et al. 1998). Changes in food and feedinghabits of sharks during growth are reflected in theiraccessibility to prey available or in the improvedcapacity of larger sharks to capture different prey(Graeber 1974; Weihs et al. 1981; Stillwell andKohler 1982; Lowe et al. 1996). Shifts in the TL ofshark species with growth have not been quantifieduntil the present. The current study represents thefirst evidence in this sense, revealing importantchanges of S. guggenheim TL throughout theirgrowth. When the population was split into lengthgroups (based on diet composition), the smallestsharks (23–60 cm LT) were identified as secondaryconsumers (TL<4) whereas medium sharks (61–80 cm LT) and largest sharks (81–91 cm LT) wereplaced as tertiary consumers (TL>4). The increase indiet diversity and the shifts in type, size, and TL of

prey ingested by S. guggenheim place this predator athigher trophic levels as it grows. However, S.guggenheim (TL=3.90, current study) cannot be con-sidered a top predator because it is, in turn, preyed uponby other larger sharks such as Carcharhinus brachyurus(Günther) (TL=4.2, Cortés 1999), Carcharias taurusRafinesque (TL=4.4, Cortés 1999) (Lucifora 2003) andNotorynchus cepedianus (Péron) (TL=4.7, Cortés1999) (Lucifora et al. 2005). According with Cortés(1999) the trophic levels of sharks ranged from 3.1 inStegostoma fasciatum (Hermann) to 4.7 in N.cepedianus. Thus, as a species S. guggenheim is asecondary consumer (TL<4) and their predators aretertiary consumers (TL>4) in the food webs in whichthese sharks occur. S. guggenheim TL estimated by thecurrent study agrees well with the values reported byCortés (1999) for six other species of Squatina.However, it’s important to note that the current studywas population-based, which allowed a more preciseestimate of TL since it was supported by the observedpopulation length frequency distribution (sample size=1280 sharks; unpublished data).

The TL of the whole population should beconsidered as a population “health” indicator since itsummarizes changes in population dynamics, preyavailability and predator–prey interactions. Then, ifthe whole population TL of S. guggenheim was lessthan 3.90 (as reference point) this will be indicative ofa decrease of the adult fraction, or an increase in thejuvenile fraction of the population or/and a change inthe environmental prey availability. Thus, the periodicestimation of the whole population TL would be auseful indicator to identify changes in the populationsize structure brought about by fishing, naturalphenomena or other causes, which could impact thepredation pressure exerted by S. guggenheim, andwith it the consequent modification of the trophicflow between components of the coastal marinecommunity. The trophic implications of removingthis predator from the coastal marine ecosystem of theAUCFZ remain unknown.

Trophic models have been long recognized asuseful computational tools to elucidate the trophicposition of the species into the food web and even toidentify the species trophic roles within a marineecosystem (e.g. Christensen and Pauly 1992; Walterset al. 1997; Cury et al. 2000; Pauly et al. 2000;Christensen and Walters 2004). However, trophicmodels do not usually split top predators, such as

Environ Biol Fish (2009) 84:41–52 49

sharks or marine mammals, according with develop-mental stages or length groups (e.g. Stevens et al.2000). The evidence presented here indicates that, inorder to obtain a more realistic representation of thestructure and dynamics of marine food webs, trophicmodels necessarily should consider changes in thetrophic level that occur during the growth of sharks.

It is evident that a single method of stomachanalysis cannot give a complete picture of dietcomposition. In order to glean the maximum amountof information from the material, the prudent inves-tigator should employ at least one method measuringthe amount, and one measuring the bulk of foodmaterial present (Hyslop 1980). This study did notuse volumetric or gravimetric methods to measure thebulk of the food material. Nevertheless, it is consid-ered that the two methods employed here to recordthe amount of food material (i.e. occurrence andnumerical methods) are good approximations in orderto study the diet composition of S. guggenheim,taking into account that: (a) is a carnivorous predatorthat swallows its prey whole and, (b) on boardidentification of prey items from stomach contentswas feasible.

Summary

Changes in the trophic ecology of S. guggenheimwere detected with its growth. First, shifts in the type,size, life habits and TL of prey ingested wereidentified. Second, the size range of prey expandedwith increasing predator body length. Third, theincrease of mean prey weight was positively correlat-ed with the increase of S. guggenheim weight. Lastly,switches of S. guggenheim TL through its growthwere quantified. Interactions between prey (type, size,TL) and predator (length, weight) characteristicsexplain the increase in S. guggenheim TL as thispredator grows larger.

Acknowledgments The authors thank DINARA (Uruguay)for permitting us to generate the database used in this researchthrough the participation of R.V.S and A.C.M. in regularMicropogonias furnieri and Merluccius hubbsi evaluationcruises. We also acknowledge the crew of the RV “Aldebarán”for their assistance. B. Yannicelli, E. Cortés and L. Luciforaprovided useful comments to improve the manuscript. TheGerman Academic Exchange Service (Deutscher Akademisher

Austausch Dienst, DAAD) funded R.V.S. (A/99/14455) and A.C.M. (A/01/17601) through scholarships to conduct graduatestudies at the University of Concepción.

References

Acha EM, Mianzan HW, Guerrero RA, Favero M, Bava J(2004) Marine fronts at the continental shelves of australSouth America: physical and ecological processes. J MarSyst 44:83–105 doi:10.1016/j.jmarsys.2003.09.005

Assis CA (1996) A generalized index for stomach contentsanalysis in fish. Sci Mar 60:385–389

Bakun A, Parrish RH (1991) Comparative studies of coastalpelagic fish reproductive habitats: the anchovy (Engraulisanchoita) of the southwestern Atlantic. ICES J Mar Sci48:343–361 doi:10.1093/icesjms/48.3.343

Bazzino G, Quiñones RA (1999) Características biológicas yecológicas del calamar Illex argentinus (Cephalopoda,Ommastrephidae) relevantes para el ordenamiento de supesquería en el Atlántico Sudoccidental. Gayana (Zool)63:87–100

Boschi EE, Fischbach CE, Iorio MI (1992) Catálogo ilustradode los crustáceos estomatópodos y decápodos marinos deArgentina. Frente Marit 10:1–92

Brewer DT, Blaber SJM, Salini JP, Farmer MJ (1995) Feedingecology of predatory fishes from Groote Eylandt in theGulf of Carpentaria, Australia, with special reference topredation on penaeid prawns. Estuar Coast Shelf Sci40:577–600 doi:10.1006/ecss.1995.0039

Cherel Y, Duhamel G (2004) Antarctic jaws: cephalopod preyof sharks in Kerguelen waters. Deep-Sea Res (1 OceanogrRes Pap) 51:17–31

Chiaramonte GE (1998) Shark fisheries in Argentina. MarFreshw Res 49:601–609 doi:10.1071/MF97136

Christensen V, Pauly D (1992) ECOPATH II—a software forbalancing steady-state ecosystem models and calculatingnetwork characteristics. Ecol Model 61:169–185doi:10.1016/0304-3800(92)90016-8

Christensen V, Walters C (2004) Ecopath with Ecosim:methods, capabilities and limitations. Ecol Model172:109–139 doi:10.1016/j.ecolmodel.2003.09.003

Clarke KL (1993) Non-parametric multivariate analyses ofchange in community structure. Aust J Ecol 18:117–143doi:10.1111/j.1442-9993.1993.tb00438.x

Clarke KL, Green RH (1988) Statistical design and analysis fora “biological effects” study. Mar Ecol Prog Ser 46:213–226 doi:10.3354/meps046213

Colonello JH, Lucifora LO, Massa AM (2007) Reproduction ofthe angular angel shark (Squatina guggenheim): geograph-ic differences, reproductive cycle, and sexual dimorphism.ICES J Mar Sci 64:131–140

Cortés E (1999) Standardized diet compositions and trophiclevels of sharks. ICES J Mar Sci 56:707–717 doi:10.1006/jmsc.1999.0489

Cortés E, Manire Ch A, Hueter RE (1996) Diet, feeding habits,and diel feeding chronology of the bonnethead shark,Sphyrna tiburo, in southwest Florida. Bull Mar Sci58:353–367

50 Environ Biol Fish (2009) 84:41–52

Cury P, Bakun A, Crawford RJM, Jarre A, Quiñones RA,Shannon LJ et al (2000) Small pelagics in upwellingsystems: patterns of interaction and structural changes in‘‘wasp-waist’’ ecosystems. ICES J Mar Sci 57:603–618doi:10.1006/jmsc.2000.0712

Ellis JR, Pawson MG, Shackley SE (1996) The comparativefeeding ecology of six species of shark and four species ofray (Elasmobranchii) in the north-east Atlantic. J Mar BiolAssoc UK 76:89–106

Froese R, Pauly D (eds) (2006) FishBase. http://www.fishbase.org. Cited Dec 2006

George EL, Hadley WF (1979) Food and habitat partitioningbetween rock bass (Ambloplites rupestris) and smallmouthbass (Micropterus dolomieui) young of the year. Trans AmFish Soc 108:253–261 doi:10.1577/1548-8659(1979)108<253:FAHPBR>2.0.CO;2

Graeber RC (1974) Food intake patterns in captive juvenilelemon sharks, Negaprion brevirostris. Copeia 1974:554–556 doi:10.2307/1442559

Guerrero RA, Acha EM, Framiñán MB, Lasta CA (1997)Physical oceanography of the Rio de la Plata Estuary,Argentina. Cont Shelf Res 17:727–742 doi:10.1016/S0278-4343(96)00061-1

Hyslop EJ (1980) Stomach content analysis: a review ofmethods and their application. J Fish Biol 17:411–429doi:10.1111/j.1095-8649.1980.tb02775.x

Karpouzi VS, Stergiou KI (2003) The relationships betweenmouth size and shape and body length for 18 species ofmarine fishes and their trophic implications. J Fish Biol62:1353–1365 doi:10.1046/j.1095-8649.2003.00118.x

Lowe CG, Wetherbee BM, Crow GL, Tester AL (1996)Ontogenetic dietary shifts and feeding behaviour of thetiger shark, Galeocerdo cuvier, in Hawaiian waters.Environ Biol Fisches 47:203–211 doi:10.1007/BF00005044

Lucifora LO (2003) Ecología y conservación de los grandestiburones costeros de Bahía Anegada, Provincia de BuenosAires, Argentina. Ph.D. thesis. Universidad de Mar delPlata, Argentina p 410

Lucifora LO, Menni RC, Escalante A (2005) Reproduction,abundance and feeding habits of the broadnose sevengillshark, Notorynchus cepedianus, in north Patagonia,Argentina. Mar Ecol Prog Ser 289:237–244 doi:10.3354/meps289237

Lyle JM (1983) Food and feeding habits of the lesser spotteddogfish, Scyliorhinus canicula (L.), in Isle of Man waters.J Fish Biol 23:725–737 doi:10.1111/j.1095-8649.1983.tb02950.x

Menni RC, Ringuelet RA, Aramburu RH (1984) Peces marinosde la Argentina y Uruguay. Hemisferio Sur, Buenos Aires,p 369

Norbis W, Lorenzo I, Torres GJ (1999) Intra-annual growthvariations of young-of-the-year hake (Merluccius hubbsi)of the Uruguayan continental shelf based on otolithanalysis. Fish Res 44:129–137 doi:10.1016/S0165-7836(99)00059-4

Olson DB, Podestá GP, Evans RH, Brown OB (1988) Temporalvariations in the separation of Brazil and MalvinasCurrents. Deep-Sea Res 35:1971–1990 doi:10.1016/0198-0149(88)90120-3

Paesch L, Meneses P (1999) La pesquería de elasmobranquiosen la Zona Común de Pesca Argentino–Uruguaya. In: ReyM, Arena G (eds) Estudios realizados sobre los elasmo-branquios dentro del Río de la Plata y la Zona Común dePesca Argentino–Uruguaya. INAPE–PNUD URU/92/003,Montevideo, 1–3

Pauly D, Trites AW, Capuli E, Christensen V (1998) Dietcomposition and trophic levels of marine mammals. ICESJ Mar Sci 55:467–481 doi:10.1006/jmsc.1997.0280

Pauly D, Christensen V, Walters C (2000) Ecopath, Ecosim,and Ecospace as tools for evaluating ecosystem impact ofFisheries. ICES J Mar Sci 57:697–706 doi:10.1006/jmsc.2000.0726

Platell ME, Potter IC, Clarke KR (1998) Resource partitioningby four species of elasmobranchs (Batoidea: Urolophidae)in coastal waters of temperate Australia. Mar Biol (Berl)131:719–734 doi:10.1007/s002270050363

Scharf FS, Juanes F, Rountree RA (2000) Predator size–preysize relationships of marine fish predators: interspecificvariation and effects of ontogeny and body size on trophicniche breadth. Mar Ecol Prog Ser 208:229–248doi:10.3354/meps208229

Schefer LN, Platell ME, Valesini FJ, Potter IC (2002)Comparisons between the influence of habitat type, seasonand body size on the dietary compositions of fish speciesin nearshore marine waters. J Exp Mar Biol Ecol 278:67–92 doi:10.1016/S0022-0981(02)00337-4

Sea Around Us (2006) A global database on marine fisheriesand ecosystems. http://www.seaaroundus.org. Cited Dec2006

Stephens DW, Krebs JR (1986) Foraging theory. PrincetonUniversity Press, Princeton, p 247

Stergiou KI, Karpouzi VS (2002) Feeding habits and trophiclevels of Mediterranean fish. Rev Fish Biol Fish 11:217–254 doi:10.1023/A:1020556722822

Stevens JD, Bonfil R, Dulvy NK, Walker PA (2000) Theeffects of fishing on sharks, rays and chimaeras(Chondrichthians), and the implications for marineecosystems. ICES J Mar Sci 57:476–494 doi:10.1006/jmsc. 2000.0724

Stillwell CE, Kohler NE (1982) Food, feeding habits, andestimates of daily ration of the shortfin mako (Isurusoxyrinchus) in the northwest Atlantic. Can J Fish AquatSci 39:407–414

Villwock de ML, Vooren CM (2003) Catch and effort ofdemersal elasmobranchs in south Brazil from 1975 to1997. Frente Marit 19:217–231

Vögler R, Milessi AC, Quiñones RA (2003) Trophic ecology ofSquatina guggenheim on the continental shelf off Uruguayand northern Argentina. J Fish Biol 62:1254–1267doi:10.1046/j.1095-8649.2003.00105.x

Vögler R, Milessi AC, Quiñones RA (2008) Influence ofenvironmental variables on the distribution of Squatinaguggenheim (Chondrichthyes, Squatinidae) in the Argen-tine–Uruguayan Common Fishing Zone. Fish Res 91:212-221 doi:10.1016/j.fishres.2007.11.028

Walters C, Christensen V, Pauly D (1997) Structuring dynamicmodel of exploited ecosystem from trophic mass-balanceassessments. Rev Fish Biol Fish 7:139–172 doi:10.1023/A:1018479526149

Environ Biol Fish (2009) 84:41–52 51

Weihs D, Keyes RS, Stalls DM (1981) Voluntary swimmingspeeds of two species of large carcharhinid sharks. Copeia1981:219–222 doi:10.2307/1444062

Werner EE, Gilliam JF (1984) The ontogenetic niche and speciesinteractions in size-structured populations. Annu Rev EcolSyst 15:393–425 doi:10.1146/annurev.es.15.110184.002141

Wetherbee BM, Cortés E (2004) Food consumption and feedinghabits. In: Carrier JC,Musick JA, HeithausMR (eds) Biologyof sharks and their relatives. CRC, Florida, pp 223–244

White WT, Platell ME, Potter IC (2004) Comparisons betweenthe diets of four abundant species of elasmobranchs in asubtropical embayment: implications for resource parti-tioning. Mar Biol (Berl) 144:439–448 doi:10.1007/s00227-003-1218-1

Wood JB, Day C (eds) (2006) CephBase. http://www.cephbase.utmb.edu. Cited Dec 2006

Zar JH (1999) Biostatistical analysis, 4thth edn. Prentice Hall,New Jersey, p 663

52 Environ Biol Fish (2009) 84:41–52