Diets and Bathymetric Distributions of Two Bathyal Sharks of the Catalan Deep Sea (Western Mediterranean)

I MARINE ECOLOGY PROGRESS SERIES Vol. 82: 21-30.1992 Published May 14 l Mar. Ecol. Prog. Ser. l

Diets and bathymetric distributions of two bathyal sharks of the Catalan deep sea (western Mediterranean)

' Unidad de Zoologia. Departamento de Biologia Animal. Universidad Autonoma de Barcelona, Bellaterra. E-08193 Barcelona, Spain Instituto de Ciencias del Mar-CSIC, Paseo Nacional s/n, E-08039 Barcelona, Spain

ABSTRACT: Bathymetric distribution of Centroscymnus coelolepis has been studied. This species was restricted to the lower slope (1419 to 2251 m), where it was the only dbundant shark. Galeus rnelastornus was abundant below 1000 m but became rare between 1400 dnd 1600 m, so that there is almost no bathymetric overlap with C. coelolepis. Greatest abundance of G. melastomus in the Catalan Sea was found in the upper slope. Diets of both species showed a very low overlap which is mainly attributed to the dietary specialization of C. coelolepis. Its diet is almost exclusively based on cephalopods. The diet of G. melastomus has been analyzed at depths below 1000 m. Its diet is more diverse than that of C. coelolepis and cephalopods are not preferential prey. The upper and middle slope specimens of G. melastomus have a different diet; this difference may be the consequence of a change in available resources. Finally the trophic position of Etrnopterus spinax, the third most abundant shark below 1000 m, seems to be more similar to that of C. coelolepis. These results suggest that a possible competition is more likely between C. coelolepis and E. spinax than between C. coelolepis and G. melastomus. This could explain the bathymetric displacement of C. coelolepis in the western Mediterranean as a result of an interaction between species of higher trophic levels.

INTRODUCTION (Capape & Zaouali 1976, 1977, Relini Orsi & Wurtz 1977, MacPherson 1980, Mattson 1981, Mauchline Information on the distribution and biology of & Gordon 1983, among others). Recently, Golani bathyal sharks in the deep Mediterranean Sea (below (1986/87) found this species below 1000 m, off the a depth of 1000 m) is limited and fragmentary. coast of Israel, and reported preliminary data on its The bathymetric distribution of Centroscymnus diet at these depths. coelolepis Bocage & Capello, 1864, the only abundant Diet analysis allows the study of the trophic overlap species at great depths in the Mediterranean Sea, among species within a community. This parameter is has been partially determined. Subsequent to the essential in determining the intensity of the inter- synthesis made by Grey (1956), the published data specific interactions (i.e. in marine fish con~munities: were incomplete (Tortonese 1956, Torchio & Michel- MacPherson 1979, 1981, Blaber & Bulman 1987, angeli 1971, Allue et al. 1985, Della Croce et al. 1988). among others). For deep-sea sharks, the available data The trophic habits of this species have been studied in on trophic overlap are scarce. Mauchline & Gordon some areas of the Atlantic and Pacific Oceans (Clarke (1983) discussed some preliminary data at Rockall & Merrett 1972, Marshal1 & Merrett 1977, Sedberry & Trough (NW Ireland); this aspect was also studied Musick 1978, Mauchline & Gordon 1983, Yano & by MacPherson (1980) in the upper slope of the Tanaka 1988), but they are completely unknown in the Catalan Sea. Mediterranean. Bathymetric distribution, trophic habits and trophic There is plentiful information about Galeus mela- interactions of Centroscymnus coelolepis and Galeus stomus Rafinesque, 1810, although it is usually limited melastomus are presented in this work. These aspects to commercial grounds, between 200 and 700 m are analyzed within groups of immature and adult

O Inter-Research/Printed in Germany Mar. Ecol. Prog. Ser. 82: 21-30, 1992

sizes. The diet of G. melastomus is descnbed and com- stomachs, and V = percentage by volume of the food pared in 2 bathymetric levels which correspond to the item in the stomachs (Hureau 1970). In our study we upper and middle slope respectively. Finally, we used weight (W)instead of volume (V).This index has discuss the possible competitive interactions and the been expressed as %IRI = (IRI/EIRI) X 100. Prey were habitat segregation among the most common sharks sorted in decreasing order according to their %IRI con- which inhabit the slope below l000 m. tribution, and then cumulative %IRI was calculated. Prey were classified into the 3 following categories according to their contribution to the cumulative %IRI MATERIALS AND METHODS (Rossechi & Nouaze 1987): preferential prey (those whose cumulative %IRI attains at least 50 % of the All material was collected from the continental slope total %lRI); secondary prey (those whose cumulative of the Catalan Sea (western Mediterranean, 38'45' to %IRI, when added to that of preferential prey, attains 42"00' N, 01' 30' to 02O 50' E),during cruises carried out at least 75 "/o of the total %IRI); and accidental prey within the framework of the BATIMAR, ABISMAR and (the remaining prey). PONT'9O projects (1988 to 3.990). The specimens were Diet was studied according to the different size- captured with a semi-balloon otter trawl (OTSB14) (cf. classes in both species. We considered 2 size-groups Merrett & Marshall 1981) and with longlines. Speci- which generally corresponded with the immature mens of Galeus melastomus from the upper slope were specimens (total length <40 cm) and mature speci- also obtained from commercial trawls. mens (total length 240 cm). The maturity size of To define the bathymetric distribution below a depth Galeus melastomus was previously determined in the of 1000 m, we used records of 76 specimens of Galeus study area by MacPherson (1980). Maturity length melastomus and 62 of Centroscj~mnus coelolepis, of Centroscymnus coelolepis was determined from captured exclusively by the OTSB14. We divided the personal observations. slope sampled (984 to 2251 m) into 6 equivalent bathy- The diet of Galeus melastomus was determined for 2 metric zones of 200 m each, and calculated the abun- different bathymetric strata (371 to 667 m and 984 to dance of the species in each zone Prior to the calcula- 1584 m), which approximately correspond to the upper tions, data were standardized to number per hour of and middle slope. All of Centroscymnus coelolepis trawling. In each zone, the number of individuals per specimens were captured in the lower slope (1729 to hour was determined as the arithmetic mean of the 2251 m). total number of samples taken within these intervals. We determined the cumulative trophic diversity A total of 86 specimens of Centroscymnus coelolepis curves in the same way as Mauchline & Gordon (1985). and 99 of Galeus melastomus were examined to deter- The Shannon-Wiener index (Shannon & Weaver 1949) mine their diet. Moreover, to compare their trophic was calculated on the abundance values of prey for habits, a few specimens of Etmopterus spinax L., 1759 each specimen. The asymptotic stabilization of the collected with the OTSB14 below 1000 m were ana- curves indicates the minimum number of stomachs lyzed. These data were not included in the results that have to be analyzed in order to get reliable because of the small number of individuals studied. enough results. Immediately after capture, specimens were fixed in Degree of overlap in the diet of both species, by 10 % formalin. Once in the laboratory, they were different sizes and bathymetric levels, was determined measured and dissected for an.alyis of stomach content. using the quantitative Schoener index (Schoener 1974) The food items were identified to the lowest tax- and the semi-quantitative Spearman correlation onomic level possible. We registered their number and coefficient (Sokal & Rohlf 1969). Trophic diversity weight to the nearest 0.1 mg, after drying them with (H') was calculated using the %IRI values and the blotting-paper to remove surface moisture. In the case Shannon-Wiener index. of the Foraminifera, we always assigned an abundance of 1 because they were passively ingested (Lagardere 1977). RESULTS The quantitative importance of every prey group in the diets of both species was determined by the Index Bathymetric distributions of Relative Importance (Pinkas et al. 1971), which was defined as: Centroscymnus coelolepis

The bathymetric preferences of Centroscymnus where F = frequency of occurrence of the food item; coelolepis in the Catalan Sea (which has a maximum N = numerical percentage of a food item in the depth of about 2400 m) are represented in Fig. 1 This Carrasson et al.: Diets and distributions of two bathyal sharks 23

more data would be required to confirm this. Adult males and females (length 240 cm) were found between 1700 and 1900 m, as were immature males and females (20 to 30 cm). The largest specimens of both sexes were not present at the greatest depths. Below a depth of 2000 m sharks of intermediate size (30 to 40 cm) prevailed.

Galeus rnelastornus

Depth (m) In the western Mediterranean, Galeus melastomus is m G. melastomus C coelolepis the most abundant demersal shark on the upper and middle slope (Relini Orsi & Wurtz 1977, Allue 1985, Fig. 1 Galeus melastomus. Centroscymnus coelolepis. Bathy - Bauchot l98fa). metric distribution of 2 bathyal sharks in the Catalan Sea In our study, we found that Galeus melastomus was below a depth of 1000 m abundant from 1000 m to as deep as 1400 m. Below that depth, numbers declined very sharply and it species is found almost exclusively on the lower slope, became rare (Fig. 1). where it is the only common shark; the upper limit is Bathymetric distribution of Galeus melastomus by about 1419 m, although it is a very rare species at size and sex is shown in Fig 2. At the limit of the depths less than 1600 m. The complete absence of this bathymetric range adult males disappear and only shark in the upper and middle slope is corroborated by immature males of 10 to 20 cm total length are the absence of published records in an area where found. On the other hand, immature and adult intensive sampling of the demersal fauna has been females of about 10 to 61 cm are present at the greatest carried out. depths. Fig. 2 represents the bathymetric distribution of all specimens by size and sex. There is a possible trend in Diet sexual and ontogenetic segregation by depth although Cen troscymnus coelolepis

Table 1 shows the results of stomach content analysis for 86 specimens. The high percentage of empty stomachs (41.2%) is an obstacle to conclusions in the case of mature individuals (Fig. 3). In general the diet consisted almost exclusively of cephalopods, decapods and fishes. This low diversity can be seen in Table 2c. The relative in~portanceof the different prey groups is shown in Fig. 4. It emphasizes the dominant role of the cephalopods, in which the combination of frequency and great abundance together with their high weight results in a %IRI of 87.2. Decapods and fishes can be considered as casual prey groups. The contribution of the remaining prey was minimal (%IRI 50.3). Important ontogenic changes in the diet are evident in Table 1. The low value of overlap between both size- groups (27.6) corroborates this (Table 2b). In both size-groups, cephalopods were always the DEPTH (m) preferred prey. Natantia decapods made an important contribution to the diet of the immatures. There was an A G. melastornus c? C. coelolepis c? obvious preference for the caridean Acanthephyra A G. rnelastomus Q o C. coelolepis Q eximia (%IRI = 12.4),which is by far the most abundant decapod in the Mediterranean deep sea (Cartes 1991). Fig. 2. Galeus melastomus, Centroscymnus coelolepis. Bathy- metr~cdistribution by slzes and sex of 2 bathyal sharks in the There was a total absence of decapods in the adult Catalan Sea below a depth of 1000 m diet. 24 Mar. Ecol. Prog. Ser. 82: 21-30, 1992

Table 1 Centroscymnus coelolepis. Qualitative and quantitative composition of diet by size-classes. F: frequency; N: percentage by number; W: percentage by weight

Foraminifera Planktonic Foraminifera Decapoda Natantia Acanthephyra eximia Acanthephyra pelagica Decapoda unid. Cephalopoda Heteroteuthis dispar Histhioteuthis sp. Todarodes sagittatus Ommastrephes bartrami Ommastrephidae ~lnid. Theuthoidea unid. Cephalopoda unid. Pisces Osteichthyes Chauliodus sloani Bathyptero~smediterraneus Osteichthyes unid. Cetacean blubber Plastic Unidentified tissues

No. fish examined % Empty stomachs No. prey items (average) Total weight (average)

Galeus melastomus determined with sufficient accuracy (Fig. 3). Three prey groups (Natantia decapods, fishes and cephalopods) Table 3 shows the results of the stomach analysis of mainly dominated the diet on the middle slope 99 individuals. The percentage of empty stomachs was (Fig. 4C). On the upper slope, euphausiids and Macrura very low (10.2 % in the middle slope and null in the decapods were also seen. Differences were also noted upper slope). In all size-groups we studied, diets were between the size-classes in each bathymetric zone

MS-a

Fig. 3. Galeus melastomus. Centroscyrnnus coelolepis. Cumulative trophic diversity curves for different size and bathymetric

groups for (m) G melastomus and (0) C. 0 5 10 15 20 25 3 0 35 40 coelolepis. Us: upper slope; MS m~ddle number of fish slope. Ls: lower slope; a: adults; i: lmmatures Carrasson et al.: Diets and distributions of two bathyal sharks 25

Table 2. Centroscymnus coelolepis, Caleus melastomus. (a) Spearman correlation coefficient between the different groups established ('significant at pS005). (b) Overlap Schoener index. (c) Trophic diversity (H').All calculations were made using %IRI. CCI: Immatures of C. coelolepis; CCA: adults of C. coelolepis; GMIU: immatures of G. melastomus (upper slope);GMAU: adults of G. melastomus (upper slope); GMIM: immatures of C. melastomus (middle slope); GMAM: adults of G. melastomus (middle slope)

a. Spearrnan correlation CCA GMIU GMAU GMIM GMAM

CC1 0.08 -0.56' -0.35 -0.05 0.30 CCA - -0.46' -0.30 -0.16 0.02 GMIU - 0.47 0.03 -0.12 GMAU - - 0.05 -0.10 GMIM - - - - 0.14

b. Overlap Schoener index CCA GMIU GMAU GMIM GMAM

CC1 27.6 5.7 5.8 28.6 23.6 CCA - 9.5 21.6 34.8 31.3 GMIU - - 23.6 20.5 16.1 GMAU - - 58 9 43.4 GMIM - - 51.4

c. Trophic diversity H'

CC1 1.34 CCA 2.48 GMlU 1.78 GMAU 2.57 GMIM 2.48 GMAM 3.47

Upper slope. We identified 45 prey from a total of 60 stomachs. Natantia decapods and euphausiids were the preferred prey, fishes and Macrura decapods were secondary and cephalopods and the rest were accidental (Fig. 4B). At the species level Calocaris Fig. 4. Relative importance of prey groups to (A) Centroscym- nus coelolepis, (B) Galeus melastomus (upper slope), and rnacandreae and Pasiphaea multidentata were the (C) Galeus melastomus (middle slope). Areas of rectangles most exploited prey; both species are predominant are proportionate to the IRI. F: Frequency in O/oof a prey group elements of the benthic and benthopelagic fauna (width of rectangle);N: percentage by number of a prey group; in the upper slope of the Catalan Sea (Abello et al. W: percentage by weight of a prey group; cep: Cephalopoda; nat. Decapoda Natantia, PIS: Pisces; for: Foraminifera; amp: 1988). Amphipoda; eup: Euphaus~acea;mac: Decapoda Macrura; There were important differences between the diets rep: Decapoda Reptantia of immature and adult specimens. These differences are shown by low values in the overlap (Table 2b). Calocarismacandreae, which is the preferential prey Middle slope. 43 prey items were identified from a of the immatures (%IRI = 62.8),is only a casual prey in total of 39 stomachs examined. In the large size-class, the adults (%IRI = 4.1). The importance of euphausiids diversity was low because Natantia decapods, fishes is also considerably reduced. On the other hand fishes and cephalopods were preferred prey (%IRI= 97.7). and Pasiphaea rnultidentataare more common in the Pasiphaea rnultidentatawas the most abundant prey adults. The cephalopods, which hardly appear in the species; its weight was less than other prey items such diet of the immatures, are a secondary prey in the as Todarodessagittatus or Mora rnoro,but it was much adults. more frequently captured. W Wil N L" U ", - I l l m l l % l S l l I l V" l ""l I l l z", z m WW~OIWWN m W- WW WW

P P O . . . I ~lplw4- I 1~"~l10-op .lad I I I I . . I . .l I I . I I . . . . . I I I .g P m cl NN -WW IOVIW WO W W + . . " 1 ;l~l~l l ."l '"-0" "1 z; l l l l . . I "l l '0 l l 0 '0 . " ' l 21 1 -g ON +N+m N +. + +

N N I I 11111 00 -, I I ,*I I I I -.* "I I m I I ~ooo01 -pro- I I I I, I I I I I I I I W- e P em o o WW~W WWWWP

- UN- - 4 11 i I Ic"L"c"S 11 -IIIc~w-u~~+o~~IIvI III~IIIII 11' . I) 11 ::'I: (DWWW m w-mw mm W W

a I l . I l 00"" I I - l l l %Gm0 l 2 l l l, ,111 ,1111 ,l 11 Ill l I 11 11 WWWpoO1g W W'dVI + VINOW Carrasson et a1 : Diets and distribut~ons of two bathyal sharks 27

Table 3 (continued)

Cephalapodd Heleroreulhls d~spar Chrroteuthisverdnyr Gdlir~~rthisarmala Histhioteuthrs sp. Todarodes saglltatus Ommastrephldaeunid. Teuthoidea unid. Bathypolypu, ~ponsal~s Cephalopoda unld

Brachiopoda Gryphus sp

Tunlcata Pyrosomida Salp~dae

P~scesOsteichthyes 41 5 13.5 2.8 Chauliodus sloani --- Stomlas boa boa --- Symbolophorus verany~ --- Notolepls nssol - - - Mesopelagic fishes --- Coelorhynchus coelorhynchus - - - Antonogadusmegalokynodon - - - Mora moro --- Ostelchthyes un~d. 41 S 13.5 2.8

Scales Chlcken leg Coal Plant debr~s Remains of carton Plastic Synthetic fibres Unident~fledt~ssues

No. fish examined 4 1 19

";P Empty stomachs 0.0 0.0 No. prey items (average) 141 (3.4) 190 (10.0) Total weight (average) 18.1 (0.4) 72.4 (3.8)

Immature fish had a high diversity diet; there were DISCUSSION AND CONCLUSIONS some prey types such as hyperiid amphipods and mysids, which did not appear in the diet of the In this work, the distribution of Centroscynlnus adults. In this group we could distinguish between coelolepis has been accurately determined. This the preferred prey (fishes, %IRI = 52.3), secondary species was exclusive to the lower slope (1500 to prey (Natantia decapods, %IRI = 27.8) and the acci- 2250 m, sensu Haedrich & Merrett 1988),where it was dental prey (cephalopods, amphipods and euphau- the only abundant shark. In the rest of the western siids). In the diet of adults we can only distinguish Mediterranean Sea it presumably shows the same between the preferred prey (Natantia decapods, preferences, although extending to abyssal grounds cephalopods and fishes, with a x%IRl = 98.1) and the over 2500 m. Grey (1956) analyzed the data of 42 indi- accidental prey. In both groups, Pasiphaea multi- viduals (1350 to 2718 m) and concluded that its center dentata was the dominant prey item, but the relative of distribution was below 2000 m. importance of other prey changed from one case to The records of this species from 200 to 400 m (cf. the other. The overlap between both size groups Bauchot 198713, Fredj & Maurin 1987) are probably showed an intermediate value and there is no cor- incorrect. They may be a consequence of the extrapo- relation in the order of importance of prey items lation of the bathymetric range shown in the Atlantic (Table 2a, b). Ocean. In Atlantic waters Centroscymnus coelolepis is Mar. Ecol. Prog. Ser. 82: 21-30, 1992

common from the middle slope to the upper rise ging activity (e.g. cetacean blubber) was rare in both (Haedrich & Merrett 1988) and off the Japanese coast size-groups. Also, the maximum size which has been it is captured between 100 to 1500 m (Yano & Tanaka recorded in the Mediterranean (about 70 cm) is much 1984, 1988). Though in the western Mediterranean less than that found in the Atlantic and Pacific Oceans there is a substantial amount of information related to (Compagno 1984). Moreover, because there are some the ichthyofauna of the upper slope (e.g. see Allue ontogenetic variations in the diet, the results of this (1985) and references cited therein), C. coelolepis has work are not comparable with those obtained else- never been sited in this bathymetric zone. where. This preference for deeper waters in the Medi- For the first time, the diet of Galeus melastomus has terranean cannot be explained by hydrographic factors been analyzed at depths between 1000 and 1600 m. such as temperature or salinity, because they remain Its diet was more diverse (H' = 1.78 to 3.47) than Cen- constant (about 13 "C and 38.4 ppm, respectively) from troscymnus coelolepis. Cephalopods did not represent 200 m downwards (Salat & Cruzado 1981). A possible a large portion of the diet of G. melastomus. It reason could be a displacement of the spatial niche of consumed epibenthic and/or benthopelagic species Centroscymnus coelolepls to avoid competing against throughout its whole bathymetric range. On the upper other species with similar trophic habits such as Galeus slope, the immatures captured mainly Calocaris melastomus. macandreae and secondarily euphausiids while the Galeus melastomus was abundant below 1000 m but adults preferred the benthopelagic species (Ommast- it suddenly became rare between 1400 to 1600 m, so rephidae, mesopelagic fishes and Pasiphaea multi- that there is almost no bathymetric overlap with dentata). This tendency remained in the middle slope Cen troscymn us coelolepis. for all sizes. As with Centroscymnus coelolepis, there are some The diets of upper- and middle-slope specimens differences in the bathymetric distribution of Galeus were different and may be the consequence of a melastomus in the Mediterranean Sea and in the change in available resources. The small contribution Atlantic Ocean. According to the literature this species of Calocans macandreae to the diet of the middle slope occurs between 55 and 1000 m (Compagno 1984), immature individuals parallels the scarceness of this preferably at the 200 to 500 m level. In Mediterranean species below 1000 m (authors unpubl.). Something waters, Bauchot (1987a) and Fredj & Maurin (1987) similar occurs with the mesopelagic prey: the eu- quote the same range. Later, Golani (1986/87) reported phausiid Meganyctiphanes norvegica (the species that on the exceptional capture of 2 individuals at 1440 m, reaches the greatest depths in the Mediterranean) is off the coast of Israel. only abundant to 1000 m (Lagardere 1977). Moreover, The greatest abundance of this species in the it is accepted that 1000 m is the lower limit of the die1 Catalan Sea is found in the upper slope (Allue 1985), mesopelagic fish migration (Goodyear et al. 1972, but we can state that it is frequent to 1400 m. Below Marshal1 1979). that depth it quickly becomes rare. Its lower bathy- The scarcity of the preferred prey and the disappear- metric limit is extensive as indicated by the capture of ance of the dominant resources resulted in a diversifi- 2 females at 1739 and 1794 m respectively. In the rest cation in the diet of Galeus melastomus below 1000 m of the western Mediterranean, Galeus melastomus is in all size-classes. In general, the ontogenic differences probably common to 1400 m, and not to 1000 m as was could be attributed to the fact that most of the adults earlier believed. select large-size prey (cephalopods, Mora moro, The diets of Centroscymnus coelolepis and Galeus Geryon longipes, Pagurus alatus), rather than amphi- melastomus showed a very low overlap which was pods, mysids and euphausiids. There was also evi- mainly due to the dietary specialization of C, coelo- dence (pieces of prey too large in relation to the pre- lepis. This species almost exclusively consumes dator's size) which perhaps implies that adult Galeus cephalopods, which accounts for the low value of its melastomus hunt in groups, as Springer (1967) already trophic diversity (H' = 1.34 to 2.48). The Mediterranean suggested for the small sharks. data suggest that C. coelolepisis not a species of strong As in the case of Centroscymnus coelolepis, Galeus scavenger habits, as has been indicated in the Atlantic melastomus only occas~onally acts as a scavenger populations (Clarke & Merrett 1972, Sedberry & species. Musick 1978, Mauchline & Gordon 1983). In the Diets of these 2 sharks have been shown to be differ- Catalan Sea it is an active predator on pelagic organ- ent as revealed by both the Spearman correlation co- isms (cephalopods and, in immatures, decapod crusta- efficient and the overlap Schoener index. These values ceans), which are known to have important migratory must be carefully interpreted. The level of prey identi- habits (Foxton 1972, Roper 1972). The presence of fication (especially cephalopods and fishes) is not the some remains that could be attributable to a scaven- same in all cases. In large measure it depends on the Carrasson et al : Diets and d~str~but~onsof two bathyal sharks

advanced level of digestion. Moreover, differences personal observations on the decapod crustacean and in resource availability in the bathyn~etnc ranges fish biomass), which causes a change in the life occupied by both species undoubtedly increase such strategy of both sharks. This tendency is also noticed in low values of overlap. other aspects of their biology. Centroscymnus coelo- If Centroscymnus coelolepis and Galeus ~nelastomus lepis shows a conservative life history, which is re- had the same bathymetric distribution, it is possible flected in low fecundity and long periods of gestation that their diets would be more similar. It is interesting of the embryos (Yano & Tanaka 1988). This species to note that the adult specimens of G. melastomus that seems to be better adapted to a poorer environment inhabit greater depths (exclusively females, see Fig. 2), (lower slope) than to the upper and middle slope, and whose bathymetric range overlaps with that of which is occupied by Galeus melastomus, an oviparous C. coelolepis, show a tendency to take cephalopods as species with a higher fecundity (Capape & Zaouali their preferred prey (observations of 2 individuals 1977). captured between 1500 and 1584 m, Table 3). The trophic position of Etmopterus spinax, the third most abundant shark below 1000 m, seems to be more Acknowledgements. For thelr help In identifying food items, similar to that of Centroscyn~nuscoelolepis than of we thank Dr M. Durfort (Departament Biologla Cel.lular, Universitat de Barcelona) and F. Giro (Departament Medl Galeus melastomus. The diet of E. spinax has been Amblent, Generalitat de Catalunya); Dr D. Lloris and P. carefully studied in Mediterranean waters, but only on Rubles (Institut Ciencles del Mar ICM-Barcelona) who identi- the upper slope (Wurtz & Vacchi 1978, MacPherson fied mesopelagic fishes; F. Pages (ICM), who identified 1980). This species shows a preference for cepha- Hydrozoa; and R. Vlllanueva and Dr P. Sanchez (ICM), who Identified cephalopods. We also thank Drs J. Retana (CREAF, lopods, especially the adult specimens which inhabit Facultat de Ciencies, UAB), P. Abello (ICM) and J. D. M. the greater depths (Wurtz & Vacchi 1978). Gordon (Dunstaffnage Marine Laboratory, Oban) for the11 In this study it was not possible to obtain a sufficient comments on the manuscnpt; Mrs. Maria Codorniu for the number of samples of Etmopterus spinax because of its Engl~shtranslation, and J. M.Anguita for drawing the figures. strong pelagic habits (Grey 1956, MacPherson 1980). Thanks are due to all the members of the research programs BATIMAR, ABISMAR and PONT'9O for thelr collaboration, However, personal observations verify the size in- specially to Drs J. Rucabado, D Lloris, J Matallanas and crease with depth, and its specialization in the capture F Sarda. BATIMAR project was f~nanced by CSIC and of cephalopods. Nineteen individuals which were CAICYT (reference PAC 86-008/1D 821). ABISMAR and collected between 984 and 1488 m were exclusively PONT'9O projects were financed by CSIC. Thls study was also adults over 30 cm length. This bathymetric range coin- supported by 'Fundacio Ca~xade Barcelona Ajuts a la Recerca, 1989. Ambit Pesca' cides with the range of Galeus melastomus and does not overlap with the range of Centroscymnus coelolepis. Cephalopods were found in the stomachs of LITERATURE CITED 4 individuals (F = 100 %). This suggests a trophic Abello, P.,Valladares, F. J., Castellon. A. (1988).Analysls of the strategy which is parallel to that of C. coelolepis. structure of decapod crustacean assemblages off the Cata- These results suggest that competition is more likely lan coast (North-West Mediterranean). Mar. Blol. 98: 39-49 between Centroscymnus coelolepis and Etmopterus Allue. C. (1985). Cornposicion y estructura de la comunldad spinax than between C. coelolepis and Galeus mela- de peces demersales frente a Barcelona (Ar'ios 1980-1981). stomus. This could explain the bathymetnc movement Thalassas 3(1):57-90 Allue, R., Allue, C, Rucabado, J., Lloris, D (1985). Ichtyo- of C. coelolepis in the western Mediterranean as a faune des eaux profondes dans la Mer Catalane. Rapp. result of the vertical ranges along the depth gradient P.-v. Reun. Comm int. Explor scient. Mer Mediterr. 29(8): being more compressed by interspecific competition 111-114 between members of higher trophic levels (Rex 1977). Bauchot, M.-L. (1987a). Scyliorhinidae. In: Fischer, W., Bauchot, M.-L., Schneider, M. (eds.) Fiches FAO d'identi- Finally it is interesting to note the variations in the fication des especes pour les besoins de la peche (Revision percentage of empty stomachs, the average of prey per 1). Medlterranee et Mer Noire. Zone de peche 37. Vol. 2. stomach and the diversity of the diet. The percentage FAO, Rome of empty stomachs increases with depth, reaching a Bauchot, M.-L. (1987b). Squalidae. In: Fischer, W., Bauchot, very high value in the case of Centroscymnus coelo- M.-L., Schneider, M. (eds.) Flches FAO d'identification des especes pour les besolns de la peche (Revision 1). lepis. In the same way, the average number of prey per Mediterranee et Mer Noire. Zone de peche 37. Vol. 2. stomach decreases with depth (from 5.5 in Galeus FAO, Rome melastomus of the upper slope to 1.3 in C. coelolepis). Blaber, S. J., Bulman, C. M. (1987).Diets of fishes of the upper The trophic diversity is also less in this species. continental slope of eastern Tasmania: content, caloric values, dietary overlap and trophic relationsh~ps.Mar. In our opinion, these factors are indicative of the in- Biol. 95: 345- 357 creasing resource scarceness along the depth gradient Capape, C , Zaouali, J. (1976). Contribution a la biologie des in the western Mediterranean (Thiel 1983, Peres 1985, Scyllorhlnidae des cotes tunlsiennes. V. Galeus mela- 30 Mar. Ecol. Prog Ser. 82: 21-30, 1992

stomus Rafinesque, 1810. Reglme alimentaire. Archs Inst. durlng the day with shr~mptrawl in a west Norwegian Pasteur Tunis 53(3): 281-292 fjord. Sarsia 66: 109-127 Capape, C., Zaouali, J. (1977). Contribution a la biologie des Mauchline, J., Gordon, J D. M. (1983). Diets of sharks and Scyliorhinldae des cbtes tunisiennes. V1 Galeus mela- chimaeroids of the Rockall Trough, northeastern Atlantic stdrnus Rafinesque, 1810. Rkpartition gcographique et Ocean. Mar. Biol. 75: 269-278 bathymetrique, sexualitb, reproduction, fecondite. Cah. Mauchline, J., Gordon, J D M. (1985). Trophic divers~tyin ~iol.-~ar.18: 449-463 deep-sea fish. J. Fish Biol. 26: 527-535 Cartes, J. E. (1991).Analisis de las comunidades y estructura Merrett. N. R., Marshall, N. B. (1981). Observations on the trofica de 10s crustaceos decapodos batlales del mar ecology of deep-sea bottom living fishes collected off Catalan Ph.D. them, Univ. Polltccnica de Catalunya northwest Africa (08°-27' N).Prog. Oceanog. 9: 185-244 Clarke, M. R., Merrett, N. R. (1972).The significance of squid, Peres, J. M. (1985). History of the Medlterranean b~otaand whale and other remains from the stomachs of bottom the colonization of the depths. In: Margalef, R. (ed.) living deep-sea fish. J. mar. biol. Ass. U.K. 52: 599-603 western Mediterranean. Pergamon Press. New York Compagno, L. J. V. (1984). FAO species catalogue. Vol. 4. Pinkas, L., Oliphant, M. S., Iverson, I.L. K. (1971).Food habits Sharks of the world. An annotated and illustrated catalogue of albacore, bluefin tuna and bonito in California waters. of sharks species known to date. Part 1. Hexanchiformes to Cal~f.Fish Game Flsh. Bull 152: 1-105 Lamniformes. FAO Fish. Synop. (125) Vol. 4, part 1 Relini Orsi, L., Wurtz, M. (1977).Patterns and overlap in the Della Croce, N., Drago, N., Flocchini, G. (1988). Ricerche feeding of two selachians of bathyal fishing grounds in the biologiche e geof~slche.Campagna oceanografica N/R Ligurian sea. Rapp. P,-v Reun.Commn int. Explor scient. 'Minerva' (14-31 8.19871. Universita di Genova, Instltuto Mer Mediterr 24(5):89-93 di Scienze Ambientali Vdrine, Rapp. Tecnico 24 Rex, M. A. (1977). Zonation In deep-sea gastropods: the im- Foxton. P. (1972). Observations on the vertical distribution of portance of biological interactions to rates of zonation. In: the genus Acanthephyra (Crustacea: Decapoda) in the Keegan, B. F., Ceidigh, P. O., Boaden. P. J. S. (eds.)Biology eastern North Atlantic, with particular reference to of benthic organisms. Pergamon Press, New Yolk species of the 'purpurea' group. Proc. R. Soc. Ed~nb Roper, C. F. E. (1972). Ecology and vertlcal distribution of 73(30):301-313 Mediterranean pelagic cephalopods. In: Mediterranean Fredj, G., Maurin, C. (1987). Les poissons dans la banque des biological studies, Final Report, Part 5. Smithsonian donnees MEDIFAUNE. Application a I'etude des carac- Institut~on,Washington teristiques de Id faune icht)ologique med~terraneenne Rosecchi, E., Nouaze, Y (1987).Compai-a~son de cinq Indices Cybium 1 l(3): 21 9-299 alimentaires utilises dans l'analyse des contenus stoma- Golani, D. (1986/87). On deep-water sharks caught off the caux. Rev. Trav. Inst. Peches. marit. 49(3-4): 111-123 Mediterranean coast of Israel. Israel. J. Zool. 34(1-2):23-31 Salat. J., Cruzado. A. (1981). Masses d'eau dans la Medi- Goodyear, R. H., Zahuranec, B. J., Pugh, L. W,Gibbs, R H Jr terranee occidentale: Mer Catalane et eaux adjacentes. (1972).Ecology and vertical distribution of Mediterranean Rapp. P.-v Reun. Commn int. Explor. scient. Mer Mddi- midwater fishes. In: Medlterranean Biological Studies, terr. 27(6).201-209 Final Report, Part 3. Smithsonian Institution, Washington Schoener, T. MT. (1974). Resource partitioning in ecological Grey, M. (1956). The distribution of fishes found below a communities. Science 185: 27-39 depth of 2000 m. F~eldiana,Zool. 36(2):75-183 Sedberry, G. R., Musick, J A. (1978). Feeding strategies of Haedrlch, R. L., Merrett. N. R. (1988).Summary atlas of deep- some demersal fishes of the continental slope and rlse off living demersal flshes in the North Atlantic Basin. J. nat. the m~d-Atlanticcoast of the USA. Mar. Biol. 44. 357-375 Hist. 22: 1325-1362 Shannon, C. E., Weaver, W. (1949). The mathematical theory Hureau, J -C. (1970). Biologie comparee de quelques poissons of communication. University of Illinois Press, Urbana antarctlques (Notothenildae) Bull. Inst, oceanogr. Monaco Sokal, R. R., RohIf, F. J. (1969).Biometry. the principles and 68(1391):3-250 practice of stat~sticsin blologlcal research. W. H Freeman Lagardere, J.-P. (1977). Recherches sur la distribution ver- and Co.. San Francisco ticale et sur I'alimentation des crustaces decapodes Springer, S. (1967). Social organization of sharks populations. benthiques de la pente continentale du golfe de Gas- In: Gilbert, R. F.. Matheuson, R. F., Rall. D. P. (eds.)Sharks, cogne. Analyse des groupements carcinologiques Bull. skates and rays. John Hopkins Press, Standford Cent. 6tud. Rech. scient., B~arritz.ll(4): 367-440 Thlel. H. (1983) Meiobenthos and nanobenthos of the deep MacPherson, E. (1979). Ecological overlap between Macro- sea. In: Rowe. G. T. (ed.)Deep-sea biology. The sea. Vol. 8. urids in the western Mediterranean Sea. Mar. Biol. 53: John Wiley & Sons. New York 1.49- 159 Torchio, M., Michelangeli, M. (1971) Prima segnalazione MacPhcrson, E. (1980) Regimen allmentaire de Galeus mela- ~n acque ital~anedi uno squalidae del genero Centroscym- stomus Rafinesque, 1810, Etrnopterus splnax (L., 1758) et nus. Natura, Milano 62/3. 241-245 Scymnorhinus licha (Bonnaterrc. 1788) en Mediterranee Tortonese, E. (1956). Leptocardia, Ciclostomata, Selachli. In: occidentale. Vie Milieu 30: 139-148 Fauna d'ltalia. 2. Calderlni. Bologna MacPherson. E. (1981). Resource partition~ng in a Medi- Wurtz, M., Vacchi, M. (1978). Ricerca di cicli nittemerali terranean demersal fish community. Mar. Ecol. Prog. Ser. nell'alimentazione dl Selac~batiall Quad. Lab Tecnol. 4: 183-193 Pesca 3 (Suppl.):155-164 Marshall. N. B. (1979). Developments in deep-sea biology. Yano, K., Tanaka, S. (1984). Some biological aspects of the Blandford Press Ltd. Poole, Dorset deep sea squaloid shark Centroscymnus from Suruga Bay, Marshall, N. B., Merrett, N. R. (1977). The existence of a Japan. Bull Jpn Soc. Fish 50(2):249-256 benthopelagic fauna on the deep-sea Deep Sea Res 24 Yano, K.,Tanaka, S. (1988) Size at maturity. Reproductive (Suppl.):483-497 cycle, fecundity, and depth segregation of the deep sea Mattson, S. (1981).The food of Galeusmelastomus, Gadiculirs squaloid sharks Centroscymnus owstoni and C. coelolepis argenteus thori, Trisoplerus esmarkii, Rhinonemus cm- in Suruga Bay, Japan. Nippon Su~sanGakkaishi 54(2): bnus and Glyptocephalus cynoglossus (Pisces) caught 167-174

This article was submitted to the editor iManuscr~ptfirst received: June 12, 1991 Revised version accepted: February 4, 1992