Structure of Mediterranean 'Cnidarian Populations In

Home , Aglaophenia pluma, Eudendrium rameum, Largest organisms, Leptopsammia, Polycyathus, Polycyathus muellerae

Homage to Ramon Marga/et,· or, Why there is such p/easure in studying nature 243 (J. D. Ros & N. Prat, eds.). 1991 . Oec% gia aquatica, 10: 243-254

STRUCTURE OF 'CNIDARIAN POPULATIONS IN MEDITERRANEAN SUBLITTORAL BENTHIC COMMUNITIES AS A RESUL T OF ADAPTATION TO DIFFERENT ENVIRONMENTAL CONDITIONS

2 JOSEP-MARIA GILIl & ENRIe BALLESTEROs

.• 1 Insti t de Ciencies del Mar (CSIC). Passeig Nacional, s/n. 08039 Barcelona. Spain 2 � Centre d'Estudis Avanc;:atsde Blanes (CSIC). Camí de Santa Barbara, s/n. 17300 Blanes. Spain

Received: March 1990

SUMMARY The presence and abundance of Cnidarians, extremely cornmon in Mediterranean sublittoral benthic communities, exerts a profound effect on the entire structure of such communities, although the roles of Anthozoans and Hydrozoans differ in accordance with their different biological characteristics. Structural differences in benthic cnidarian populations were observed in three sublittoral cornmunities located within 100 m of each other yet differing considerably in light intensity and water movement. Species-area and Shannon's diversity-area curves were computed for each community from reticulate samples constituted by 18 subsamples 2 of 289 cm . Species-area curves were fitted to semilogarithmic functions, while diversity-area curves were fitted to Michaelis-Menten functions by the method of least squares. Species richness, species distribution, alpha-diversity, and pattem diversity were estimated from the fitted curves. Patch size for Anthozoans was smaller than that for Hydrozoans in all three communities. Diversity was greatest in the cornmunities with the lowest and with the highest light intensity and water flow levels, but species richness was highest in the intermediate community. The most structured community (with the most heterogeneous species distribution pattem) had the largest organisms with the longest life spans. It is suggested that the difference of water flow conditions in these three cornmunities leads to the development of different spatial pattems of individuals and colonies. The size, shape and small-scale distribution of these organisms are probably correlated to their ability of enhancing prey capture under different water flowconditions.

KEY WORDS: Cnidarians, benthos, population structure, diversity-area curves, Westem Mediterranean.

INTRODUCTION sciaphilic cornmumtIes where algal growth is limited by low light irradiance. In such Cnidarians are extremely abundant in places, diffe rent cornmunities develop in nearly all Mediterranean marine benthic relation to light intensity. Strong changes in cornmumtIes. However, the highest irradiance can occur in short spaces where number of species is found on rocky holes and caves open on rocky walls. Light sublittoral bottoms, where cnidarians make attenuation from outside to inside the caves up a substantial proportion of the biomass involves a gradient in species composition of such cornmunities (TRUE, 1970; GILI & along an horizontal axis, a gradient which ROS, 1985a). Cnidarians constitute one of is enhanced by the decrease in water the dominant group of benthic organisms in hydrodynamism along the same axis 244 GILI & BALLESTEROS

(RIEDL, 1966). This gradual change in small colonies or less aggregated species composition has led Mediterranean individuals alter the flow over the benthic biologists to define three different substratum, turbulence and pattern of communities (PÉRES & PICARD, 1964; partic1e transport (ECKMAN, 1983; ROS et al., 1985; GILI & ROS, 1985b) OKAMURA, 1984). Our hypothesis is that named (from outside to inside) the different hydrodinamical regimes in coralligenous community, semidark cave these three benthic communities, partly community and dark cave community. because of resulting differences in food According to previous observations supply, do condition the spatial distribution made in sublittoral communities (RIEDL, of the species, and therefore the population 1966; GILI et al., 1986; ZABALA et aL., structure and patch size. 1989), the water flow is intense and The structure of a community is bidirectional in the coralligenous, moderate determined by the spatial distribution of the and unidirectional at the entrance of the species within it. Studies designed to find a caves and reduced even though constant representative sampling area, based on inside the cave. This decrease in water flow species-area curves, diversity-area curves, intensity involves a parallel zooplankton and spatial variations in biomass, provide a depletion (RIEDL, 1966; M. Zabala, considerable amount of information on personal communication). community structure for ca1culating patch Passive organisms, such as hydroids and size. Thus, structure and minimal area are anthozoans, are particularly dependent on c10sely related and, even though minimal adequate degrees of water movement sampling area may vary according to the (RIEDL, 1971). Many benthic cnidarians criteria applied by different researchers are passive suspension feeders and depend (e.g., BOUDOURESQUE & BELSHER, upon water current to carry suspended food 1979; HAWKINS & HARTNOLL, 1980; across their food-capturing surfaces. The BALLESTEROS, 1986), a series of ability and distribution of benthic numerical parameters can be obtained for cnidarians is enhanced to maximize the comparing the structural complexity of volume of food-Iaden water they encounter communities. The interest in obtaining such per unit time (W AINWRIGHT & KOEHL, parameters is obvious since topics such as 1976). patch-size can be quantified and diffe rent The shape, size and spatial distribution commumt1es can easily be compared of benthic cnidarians affects the water flow (BALLESTEROS, 1986). around them. Individuals or colonies can The object of the present study is to survive in rapid or moderate flow if they examine differences in the structure of three are small, aggregate or if they adjust their benthic cnidarian populations occupying a body shape and size to surrounding objects short but sharp light-intensity and or organisms, (KOEHL, 1984). These water-flow gradient. The consequence of organisms, whose structures affect the such environmental gradients would be that water flow, operate like obstac1es that colonies and individuals would have a decrease fluid velocity and increase the different spatial pattern in each cornrnunity turbulence in their surface layer (VOGEL, which lead to different patch-sizes. 1981; LaBARBERA, 1984) and, therefore, Cnidarians make up a taxocoenosis that can enhance the possibilities of prey retention be quite representative of the community as and catch. Under fast flow conditions, a whole (WEINBERG, 1978; GILI et aL., feeding ability could be enhanced by the 1989a) and for this reason they may be a presence of neighbours with high body size good indicator of the structural pattero (SEBENS, 1984). However, under within each benthic community. moderate flow velocities, populations of CNIDARIAN POPULATION STRUCTURE IN THE MEDITERRAN EAN 245

MA TERIALS AND METHODS denudation (BOUDOURESQUE, 1971), scraping a surface consisting of 3 x 6 Samples were collected off the squares measuring 17 x 17 cm each. So, northeastem coast of the Medes Islands each sample consisted of 18 squares of 289 2 (Costa Brava, Northwestem Mediterranean, cm , making a total surface area gf 5,202 2 UTM 31TEG 15), where transition between cm (Fig. 1). The squares (subsamples) the coralligenous community and the dark were separated in the laboratory, and the cave community occurs in a reduced space hydrozoans and anthozoans were quantified (Fig. 1). in terms of: Three samples were obtained by - total biomass, measured as dry weight

102 cm

51 cm

. ...

15 Dark - cave

FIGURE l. Diagrammatic localization and configuration of the three communities studied. The reticulate rectangle abo ve the community diagrams shows the spatial distribution of samples in each community. The width and direction of arrows roughly indicates the intensity and direction of water movement in each community. 246 GILI & BALLESTEROS in grams, after drying each species at semilogarithmic coordinates. 110 Oc for 24 h; 4) The Molinier 20/5 point (NÉDÉLEC, - organic matter, i.e., ash-free dry 1979), corresponding to the point on the weight; ashes were prepared by buming species-area curve at which a 20 % increase samples for 6 h at 500 oC, in order to in surface area yields a 5 % increase in the minimize the importance of accreted number of speciés; this point can be skeletons. regarded as the minimum species patch size Two data matrices were thereby obtained . value (BALLESTEROS, 1986). for the samples, in each of which Xij was 5) Species diversity (A), the asymptotic' the biomass (or organic matter) for species value of diversity when the diversity-area i in subsample j. Species-area and curve is fitted to a Michaelis-Menten diversity-area curves were derived from function. these data matrices for total cnidarians, 6) Pattem diversity (S), the value of anthozoans, and hydrozoans. The surface area corresponding to the Calleja 3 species-area curves were fitted to a 1 . 10- point on the diversity-area curve semilogarithmic function of the form when fitted to a Michaelis-Menten function; this parameter also fumishes the area at y = a In x + b, which diversity practically stabilizes (minimal structural area; NIELL, 1974). where was the number of species and AH these parameters were used as y 2 x was the surface area in cm . The indicators in evaluating the structure of the diversity-area curves were fitted to a benthic cnidarian populations found. Michaelis-Menten function:

y = Ax / (B + x) RESULTS

were was the diversity in bits and x the A total of 39 cnidarian species were y 2 surface area in cm . recorded in the three communities (Table The following structural parameters 1). Paramuricea clavata was the dominant (BALLESTEROS, 1986) were then species in the coralligenous community, ca1culated: although other species of solitary 1) Number of species per sample. rnadreporarians, zoantharians, and 2) Species distribution (k): the value of a1cyonarians formed small but dense k in the fitted species-area curve; the colonies that contributed a substantia1 semilogarithmic equation thus obtained can portion to the total biomass of the be expressed as a power equation: community. Unlike the other communities, sorne species of hydrozoans were present, x = k.e Ay forming dense colonies that made up a significant share of the total biomass. The The parameter k is a good index for semidark-cave community presented the defining the shape of the curve, indicating largest number of species. Srnall colonies changes in slope for different patch sizes. of Hoplangia durotrix predominated in this The higher the value of k, the larger the community, although other solitary surface area required to obtain a madreporarians also had high biomass representative number of species for the values. An important number of species community. shared in the hydrozoan biornass, and 3) The parameter a, related to species colonies were not large. In the dark-cave richness, equal to the slope of the line community four madreporarian species fitting the species-area curve on the account for a great part of the total CNIDARIAN POPULATION STRUCTURE IN THE MEDITERRANEAN 247

2 TABLE 1. Biomass (n) in grams (dry weight) per m , and frequency (f) in 18 samples for all species in the three (coralligenous, semidark-cave, and dark-cave) communities considered.

Species Community

Coralligenous Semi-dark cave Dark-cave n f n f n

HYDROZOA Bougainvillia ramosa 0.0124 0.9838 5 Perigonymus repens 0.0986 l Clava multicomis 0.0880 1 Eudendrium motzkossowskae 0.0098 l 0.49 12 4 Eudendrium capillare 0.1406 2 0.4402 3 Halecium tenellum 3.5226 5 0.0408 1 Halecium labrosum 1.3380 4 0.8912 5 Clytia haemisphaerica 0.8790 4 0.4246 10 0.0064 Obelia dichotoma 1.3814 8 0.3316 12 Hebella scandens 0.0078 1 0.0394 5 Sertularella gaudichaudi 16.7862 4 0. 1028 1 Antennella secundaria 0.1624 2 0.0846 Aglaophenia octodonta 0.0158 1 Coryne pusilla 0.0060 1 Eudendrium racemosum 2.3782 11 1.1588 2 Halecium mediterraneum 0.0100 1 Phialella quadrata 0.0676 2 Laomedea pelagica 0.0368 2 Synthecium evansi 0.2806 3 Plumularia setacea 0.0400 2 Nemertesia antennina 0.0582 4 0.0080 1 Aglaophenia pluma 0.0002 1 0.0040 1 Eudendrium rameum 1.7194 3 Laomedea angulata 0.0032 l

SCYPHOZOA Stephanoscyphus eumedusoides 0.0580 6

ANTHOZOA Alcyonium acaule 8.6820 5 Parerythropodium coralloides 8.1902 12 0.1192 3 0.0098 Paramuricea clavata 266.9146 9 Epizoanthus arenaceus 0. 1 934 1 Parazoanthus axinellae 14.8068 8 Hoplangia durotrix 1.3972 1 162.5834 13 183.7603 17 Caryophyllia smithi 11.4878 9 6.0980 5 1.9984 4 Caryophyl1ia inomata 14.8224 6 41.9614 8 279.0548 7 Leptopsammia pruvoti 83.3340 12 33.9306 7 327.0448 17 Corallium rubrum 1.8490 1 58.5484 6 Comularia comucopiae 0.0354 3 Rolandia rosea 11.0832 13 0.7326 9 Corynactis viridis 6.0132 12 0. 1 672 4 Polycyathus muellerae 289.4380 3

biomass. Two of these species formed hydrozoan biomass (Table 1). dense colomes, moderate (H. durotrix) or Figure 2 sets out in more general and large (P. muelle rae) in size. Unlike the surnmarized form the percentage of species other two cornmunities, the solitary in each of the three communities according madreporarians formed facies, while just to three categories, namely, average life two species accounted for nearly the total span, biomass (dry weight), and polyp size 248 GIL! & BALLESTEROS

(taken as rnouth size). Each of these three cornrnunity the highest percentages categories has been di vided into fi ve corresponded to the lower and upper subdivisions. In the coralligenous subdivisions in all three categories; in the

CoralligMous S�midark-ca ve Dark- cav� Community Community Communi ty

L/FE SPAN

B/OMASS (dry weigh tJ

POLYP S/ZE

FIGURE 2. Percentage number of species in each of the three (coralJigenous, semidark-cave, dark-cave) communities studied in terms of life-span, biomass (dry weight), and polyp size (based on mouth size). Each category has been divided into five subdivisions. CNIDARIAN POPULATION STRUCTURE IN THE MEDITERRANEAN 249

III. T ABLE Il. Parameters calculated from the TABLE Parameters calculated from the species-area curves (the number of species, N, refecs divfrsity-area curves (S is a measure 0f pa«:h size in to the total number collected in each community; tfe cm ). -, oot significant data. Molinier 20/5 point is a measure of patch size in cm ). A . S k a N M20/5 TOTAL CNIDARIANS TOT AL CNIDARIANS Coralligenous Coralligenous 129.9 6.245 0.999 23 4976 dry weight 2.08 508 0.997 Semidark-cave 100.4 6.793 0.999 27 3846 ash-free dry weight 1.49 320 0.963 Dark-cave 113.9 4.577 0.999 17 4636 Semidark-cave dry weight 2.02 317 0.787 ANTHOZOANS ash-free dry weight 3.05 575 0.997 Coralligenous 38.4 1.857 0.993 9 1472 Dark-cave Semidark-cave 32.4 1.812 0.983 9 1240 dry weight 2.25 565 0.959 Dark-cave 42.8 2.180 0.993 10 1640 ash-free dry weight 2.47 617 0.971

HYDROZOANS ANTHOZOANS Coralligenous 176.0 4.075 0.988 14 6742 Coralligenous Semidark-cave 150.9 4.980 0.995 18 5779 dry weight 1.77 441 0.999 Dark-cave 275.7 2.398 0.985 7 10559 ash-free dry weight 1.10 228 0.911 Semidark-cave dry weight 1.89 424 0.936 semidark-cave community the percentage ash-free dry weight 2.49 565 0.997 of species in each subdivision within each Dark-cave category was approximately equal; and in dry weight 2.24 559 0.961 ash-free dry weight 2.28 571 0.%9 the dark-cave community the percentage of species became progressively larger in the HYDROZOANS upper part of the range for each category. Coralligenous The fits of the species-area curves dry weight 2.34 555 0.974 presented higher values of associated ash-free dry weight 2.46 585 0.977 k, Semidark-cave with species distribution over surface area, dry weight 3.26 790 0.998 in the coralligenous and dark-cave ash-free dry weight 3.08 742 0.998 communities (Table I1). On the other hand, Dark-cave dry weight the highest value of a, associated with ash-free dry weight species richness, was attained in the semidark-cave cornmunity. The trends for these two parameters were very similar for whereas for hydrozoans the highest value the anthozoans In the first two was for the semidark-cave community. The commumtles, whereas there were low biomass of hydrozoans in the dark-cave substantial differences for the hydrozoans: community gave rise to non-significant fits the value of a was highest in the semidark of the diversity-area curves to the cave community, while the dark-cave Michaelis-Menten function for this community yielded the highest value of k community. For biomass induding ashes, (Table II). the pattern-diversity (S) value for total The fits of the diversity-area curves gave cnidarians was highest for the dark-cave similar values for speci� diversity in all rornmunity and lowest · for the three communities. However, there were semidark-cave cornmunity. In contrast, for appreciable differences between the value ash-free biomass, the value of S increased of A for biomass as dry weight and for from the coralligenous community to the biomass as ash-free dry weight, particularly dark-cave community (Ta1'tlt III). The in the semidark cave cornrnunity. The values of S for anthozoans alone followed a value of A for anthozoans followed the similar trend to those for total cnidarians, same trend as for the total cnidarians, but for the hydrozoans the value was lower 250 GILI & BALLESTEROS in the coralligenous community both for in the coralligenous cornmunity for both ash-free biomass and biomass inc1uding anthozoans and total cnidarians. ashes. The behaviour of the two types of curve considered (see the shapes of the curves in DISCUSSION figure 3) suggests that the values of k and S fumish very similar indications of patch Patch size for total cnidarians based on size in the three cornrnunities. In any case, the species-area curves was sornewhat patch size for total cnidarians was slightly larger in the coralligenous cornrnunity than higher qualitatively in the coralligenous in the cave cornrnunities because of the cornrnunity and slightly higher distribution of the dorninant species (P. quantitatively in the dark-cave cornmunity. clavata), in the forrner cornrnunity, whose Patch size was smallest 9.r the larger and long lived colo nies affect the sernidark-cave comrnunity. For antbozoan1 s settlernent and developrnent of the other patch size was highest in the dark-cave species, which grew scattered in the basal cornrnunity, lowest in the sernidark-cave straturn. On the other hand, the lower level comrnunity. In contrast, for hydrozoans of spatial cornpetition in the cave patch size was qualitatively highest in the cornrnumtIes probably enables certain dark-cave and coralligenous cornrnunities rnadreporarians to form large colonies by but quantitatively highest in the budding as well as dense aggregations of sernidark-cave cornmunity. For ash-free srnall colonies or individuals through larval biornass, however, patch size was smallest recruitrnent, resulting in a srnaller

CORALLlGENOUS COMMUNI TY SEM IDARK-CA VE COMMUNITY

3. 3. 2.5 ��.---.----.-----. 2.0 �-I:i.--- I1-- I1----- 11 ::... •/ 11 ..____1s. 7 5 . �Q: / � 1. es Q o.

FIGURE 3. Species-area and diversity-area curves for the three (coralligenous, semidark-cave, and dark-cave) communities for each of the three groups (total cnidarians, hydrozoans, and anthozoans) considered, with CNIDARIAN POPULATION STRUCTURE IN THE MEDITERRANEAN 251

patch-size. to coexist. Both solitary anthozoan species From a quantitative point of view, the and small colonies of anthozoans and larger patch size for cnidarians in the hydrozoans contributed to this. Unlike the dark-cave and coralligenous communities, hydrozoans in the other two communities, compared with that in ·the semidark-cave the hydrozoans in the semidark-cave cornmunity, is caused by the presence of community enjoyed a greater number of larger species, which attain high biomass potential sites to colonize among the levels, in the first two. These species left anthozoan individuals and �olonies. little room for other species, which were Quantitatively, the difference in patch therefore confined to smaller patches. As a size as measured in terms of total dry consequence, in the coralligenous weight and ash-free dry weight is indicative community a high percentage of larger of the importance of the skeletal carbonates species with high biomasses coexisted with of certain anthozoans in deterrnining the many other much smaller species (Fig. 3). structural pattems in the communities. This This was still more evident in the dark-cave was quite c1ear in the case of the community, in which patch size was largest coralligenous cornmunity, in which the thanks to the patchy distribution of difference between P. clavata and all the madreporarian colonies. other species increased, and, as a result, The lowest patch size occurred in the patch size and diversity decreased, when semidark -cave community, reflecting the the skeletal weight contribution was occupation of the available space by disregarded. In contrast, in the semidark smaller species. Individuals or colonies of cave cornmunity the converse held true, these, in tum, occupied smaller areas, while in the dark-cave cornmunity there thereby allowing a larger number of species was hardly . any change. Due to the predominance of madreporarians in these last two communities, which, compared to gorgonians have high organic matter DARK-CA VE COMMUNI TY contents despite their ca1cium skeletons 11 TOTAL CNIDARIANS (GILI & ROS, 1985b), differences between • HYDROZOANS oANTHOZOANS species biomass measured as dry weight and as ash-free dry weight were minimal in the dark-cave community and decreased (i.e., diversity increased) in the semidark-cave cornmunity. Dominance of a community by a single species, termed a facies, also exerts an important effect on patch size (ROS et al., 1985). This was true in the case of P. clavata in the coralligenous community and, in part, of P. 3- muellerae in the dark-cave cornmunity. 2. /� �_!� Such dominance caused total cnidarian 2. biomass to be higher in these two /11 communities (GILI & ROS, 1985b). In /� contrast, in a study of similar communities • TOTAL CNIDARIANS off Marseilles by TRUE (1970), the highest 11 DRY WEIGHT AASH-FREE DRY WEIGHT biomass was recorded for a semidark -cave community, occupied by a facies of 1000 cffil Corallium rubrum; for this reason, patch quantitative curves for both biomass measures (dry size was quite similar to that in a weight and ash-free dry weight) used. 252 GILI & BALLESTEROS

coralligenous cornrnunity, also occupied by movement allows the zooplankton and a facies of P. clavata. oxygen concentrations near the organisms Benthic cnidarians are suspension to faH and, thus, become intolerable to most feeders and as such are dependent upon the sessile cnidarians, like hydroids or sorne external availability of food transported by octocorals, with high food requirements. In the water flow. In the context of a flow the coralligenous community again, under gradient like that existing in the three conditions of high water movement, the commumtles studied, it would seem zooplankton availabi lity increases and reasonable to expect potential food sources allows the development of a diversified to be most abundant in the coralligenous population of benthic cnidarians; large community and least abundant in the dark colonies (gorgonians) close to others of cave cornmunity. Greater food availability moderate (e.g. a1cyonaríans) or small size and greater colonizing ability would be (e.g. hydroids). However, the lower associated with a larger number of species availability of food in the dark-cave and, therefore, with a larger patch-size community would act, probably, to reduce (WRIGHT, 1983; KEOUGH, 1984). competition so that the species were able to However, the community situated in grow and occupy a large space (SEBENS, between the two extremes had the smallest 1982). patch size and the highest number of The moderate water movement In species. The largest patch size found in the semidark-cave community enables the coralligenous community is consistent with establishment of an intermediate spatial the observations by WRIGHT (1983) and distribution pattern as a consequence of the would be caused by the large area <::overed development of colonies or individuals of by P. clavata, which has a colonial habit moderate and small size. U nder these well adapted to capturing prey in high moderate flow conditions, feeding ability water flows (LEVERSEE, 1976), and a could be enhanced by the presence of certain gregarious nature that apparently neighbours, or by forming aggregations of facilitates the capture of prey in certain small and clonal colonies (ECKMAN, anthozoan species (McFADDEN, 1986). 1983; McFADDEN, 1986). Sessile cnidarians found in In consequence, as shown in figure 2, slowly-moving water, in the dark-cave the coralligenous and dark-cave community, tend to have rnorphological communities contained higher percentages and physiological adaptations which of longer-lived species that capture larger compensate for the lack of external prey items than did the semidark-cave transport (WAIN WRIGHT & KOEHL, community. In this type of species, which 1976), among them: tolerance of low have adopted more of a K strategy (sensu oxygen levels; fe eding on motile prey in MacARTHUR & WILSON, 1967; see GILI lower concentrations; aperiodical & ROS, 1985a), the level of competition appearance of potential prey. At the same with other species similar in habits and time, scleractinians in dark-caves may have biology is higher (CONNELL, 1975), a lower energy demand than other benthic which tends to favour coexistence by more cnidarians in temperate waters (SEBENS, dissimilar species, like the r-strategists or 1987). In the less accessible parts of éaves opportunistic ones (CONNELL & the amount of water movement is not SLATYER, 1977; GILI & ROS, 1985a), adequate to cover the demands for the which are smaller, shorter-lived, and more development of dense populations like specialized in the capture of smaller prey. those found in the entrance of the cave or In the semidark-cave community the lower, in the coralligenous community (RIEDL, yet nonetheless sufficient, level of food 1971). The mínimum range of water availability enabled colonies and CNIDARIAN POPULATION STRUCTURE IN THE MEDITERRANEAN 253 individuals to develop but not . to attain coralligenous cornmunity with P. clavata or large sizes. This, in turn, enabled a larger in highly food-limited environments such as number of species to coexist, but the in the dark-cave cornmunities, where patch-size would decrease because of the species with low food requirements can lack of compartmentation caused by survi ve. N evertheless, if there are few spatial-structuring species such as P. biomass-storers, or if there are none, clavata in the coralligenous cornmunity and specially adapted to a given environment, the total quantitative dominance of species richness could be increased by the Hoplangia durotrix. The qualitative appearance of opportunistic forms (e.g. dominance by hydroids and by a few hydroids) due to lower competitive levels. octocorals in the semidark -cave community AIso, patch-size decreases because of the is indicative of the ecotone character of this lack of a rigid structural pattern and low community, dominated as it is by spatial heterogeneity. r-strategists. Spatial arrangements of particular The seasonality observed in the major populations can probably be correlated with part of the Mediterranean hydroids (e.g. the ways in which certain colonies and BOERO & FRESI, 1986; LLOBET et al., individuals modify water currents (e.g., 1991) could change the spatial pattern HISCOCK, 1983) and affect the flow observed In communities such as forces and feeding currents. This suggestion semidark-caves, where these organisms reveals the importance of size, shape and play an important structural role. But their patch distribution of these animals to their great capacity for continuous growth and ability to utilize moving water, catch recruitment enhance their presence and suspension preys and successfully colonize similar distribution in this community sublittoral benthic communities. (GILI et al. , 1989b). In conc1usion, in communities in which the growth of certain biornass-storing, ACKNOWLEDGEMENTS structure-creating species is enhanced, there is a tendency towards the contagious We wish to thank Dr. J.D. Ros for his distribution of individuals of a single useful cornments and criticism of the species, resulting in larger patch-sizes. This rnanuscript and Mr. R. Sacks for helping occurs in a situation of large food with the English version. availability as is the case in the

REFERENCES

BALLESTEROS, E. 1986. Métodos de análisis Diamond, eds.): 460-490. Harvard University estructural en comunidades naturales, en particular Press. Cambridge, Mass. del fitobentos. Oecol. aquat., 8: 117-131. CONNELL, J.H. & SLATYER, R.O. 1977. BOERO, F. & FRESI, E. 1986. ZORation and Mechanisms of succession in natural comrnunities evolution of a rocky bottom hydroid community. and their role in cornmunity stability and P.5.ZN. 1: Mar. Bio!., 7: 123-1 50. organization. Am. Nat. 111: 1119-1 144. BOUDOURESQUE, C.F. 1971. Méthodes d' étude ECKMAN, J.E. 1983. Hydrodynarnic processes qualitative et quantitative du benthos (en affecting benthic recruitrnent. Limnol. Oceanogr., particulier du phytobenthos). Téthys. 3: 79- 104. 28: 241 -257. BOUDOURESQUE, C.F. & BELSHER, T. 1979. Le GILI, J.M., MURILLO J. & ROS, J.D. 1989a. The peuplement algal du port de Port-Vendres: distribution pattern of benthic Cnidarians in the recherche� sur I'aire rninimale uali ative . q t Cah. Western Mediterranean. Scient. Mar., 5 3: 1 9-35. Biol. Mar. 20: 259-269. GlLI, J.M., RIERA, T. & ZABALA, M. 1986. CONNELL, J.H. 1975. Sorne mecharusrns producing Physical and. biological gradients in a submarine structure in natural cornrnunities: a rnodel and cave on the Westem Mediterranean coast (NE evidence from field experiments. In: Ecology and Spain). Mar. Bio/., 90: 29 1 -297. Evo/ution of communities. (M. Cody & J. GILI, J.M. & ROS,lD. 1985a. Estudio cuantitativo de 254 GILI & BALLESTEROS

tres poblaciones infralitorales de cnidarios OKAMURA, B. 1 984. The effects of ambient f10w bentónicos. Inv. Pesq., 49: 323-352. velocity, colony size, and upstream colonies on the GILI, J.M. & ROS, 1.D. 1985b. Study and feeding success of Bryozoa. 1. Bugula stolonije ra Cartography of the benthic communities of Ryland, and arborescent species. 1. Exp. Mar. Biol. Medes Islands (NE Spain). P.S.z.N. I Mar. Ecol., Ecol., 83: 179-193. É E 6: 219-238. P R S, J.M. & PICARD, J. 1964. Nouveau manuel GIL!, J.M., ROS, J.D. & ROMERO, J. 1889b. The de bionomie benthique de la Mer Méditerranée. structural role of Cnidaria in Mediterranean benthic Rec. Trav. Sto mar. Endoume., 31: 5- 1 37. communities: a simulation-like approach. In: RIEDL, R. 1966. Biologie der MeereshOhlen. Paul Proceedings of the 21st European Marine Biology Parey. Berlín. Symposium (Klekowski, R., Styczynska- Jurewicz, RIEDL, R. 1971. Water movement. In: Marine E. & Falkowski, L., eds.): 279-289. Ossolineum. Ecology, 1, Environmental factors. (O. Kinne, ed.): WrocIaw. 1123-1156. Wiley. London. HA WKINS, S.1. & HARTNOLL, R.G. 1980. A study ROS, J.D., ROMERO, J, BALLESTEROS, E. & GIL!, of the small-scale relationship between species J.M. 1985. Diving in blue water. The benthos. In: number and area on a rocky shore. Estuar. Coast. Western Mediterranean (R. Margalef, ed.): Mar. Sci. , 10: 201-214. 233-295. Pergamon. London. HISCOCK, K. 1983. Water movement. In: Sublittoral SEBENS, K.P. 1982. Competition for space: growth ecology. The ecology of the shallow sublittoral rate, reproductive output and escape in size. Am. benthos. (R. Earll & D.G. Erwin, eds.): 58-96. Nat. 120: 1 89- 1 97. Clarendon. Oxford. SEBENS, K.P. 1984. Water flow and coral colony KEOUGH, MJ. 1984. Effects of patch size on the size; Inter-habitat comparisons of the octocoral abundance of sessile marine ínvertebrates. Ecology, Alcyonium siderium. Proc. Natl. Acad. Sci. USA , 65: 423-437. 81: 5473-5477. KOEHL, M.A.R. 1984. How do benthic organisms SEBENS, K.P. 1987. Coelenterata. In: Animal withstand moving water? Amer. Zool. , 24: 57-70. Energetics. 1. (TJ. Pandian & F.J. Vernberg, eds.): LaBARBERA, M. 1984. Feeding currents and particle 55-120. Academic Press. San Diego. capture mechanisms in suspension feeding animals. TRUE, M.A. 1 970. Étude quantitative de quatre Am. 2001. , 24: 71-84. peuplements sciaphiles sur substrat rocheux dans la LEVERSEE, G.J. 1976. Flow and feeding in région marseillaise. Bull. Inst. Océanogr. Monaco. fan-shaped colonies of the gorgonian coral, 1410: 1-48. Leptogorgia. Biol. Bull., 151: 344-356. VOGEL, S. 1981. Lije in moving fluids. Willard Grant. LLOBET, 1., GIL!, J.M. & HUGHES, R.G. 1991. Boston. Horizontal, vertical and seasonal distribution of WAINWRIGHT, S.A. & KOEHL, M.A.R. 1976. The epiphytic hydrozoa on the alga Halimeda tuna in nature of flow and the reaction of benthic Cnidaria the North-western Mediterranean Sea. Mar. Biol. , to it. In: Coe/enterate Ecology and Behavior. (G.O. 110: 151-159. Mackie, ed.): 5-21. Plenum. New York. MacARTHUR, R.H. & WILSON, E.O. 1967. The WEINBERG, S. 1978. The minimal area problem in theory of island biogeography. Princenton invertebrate communities of Mediterranean rocky University Press. Princenton, NJ. substrata. Mar. Biol., 49: 3-40. McFADDEN, C.S. 1986. Colony fission increases WRIGHT, D.H. 1 983. Species-energy theory: an particIe capture rates of a soft coral: advantages of extension of species-area theory. Oikos,. 41: being a small colony. J. Exp. Mar. Biol. Ecol. , 103: 496-506. 1-20. ZABALA, M., RIERA, T., GIL!, J.M., BARANGÉ, É É É Ñ N D LEC, C. 1979. tude structurale et problemes M., LOBO, A. & PE UELAS, J. 1989. Water d'échantillonage dans une phytocoenose portuaire. flow, trophic depletion, and benthic macrofauna Mem. Univ. P. et M. Curie. Marseille-Luminy impoverishment in a submarine cave from the NIELL, F.x. 1974. Les applications de l'index de Western Mediterranean. P.S.z.N. 1: Mar. Eco 1. , 10: Shannon a l'étude de la végétation intertidale. Soco 271 -287. Phycol. Fr. Bull., 19: 238-254.