original scientific paper UDK 58/59(262.3-17) LONG-TER M CHANGE S IN TH E BENTHO S O F TH E NORTHER N ADRIATI C SEA Michael STACHOWITSCH PhD, marine biologist, Institute of Zoology, University of Vienna, A-"!090 Vienna, Althanstrasse 14, E mail: @ mischkas zoo.univie. ac . at, Austria dr., morski biolog, Issstitute of Zoology, University of Vienna, A-1090 Vienna, Althanstrasse 14, E mail: @ mischkas zoo.univie. ac. at, Austria Alexander FUCHS BSc., marine biologist, Institute of Zoology, University of Vienna, A-109Q Vienna, Althanstrasse 14, Austria dipl. biolog. Institute of Zoology, University of Vienna, A-1090 Vienna, Althanstrasse 14, Austria ABSTRACT The Northern Adriatic Sea, like many shallow coastal waters, is subject to a wide range of stresses and threats. These include oxygen deficiencies, marine snow events, and fishing pressure, A mass mortality event in 1983 led to the collapse of a wide-ranging benthic community. A long-term sampling series between 1984 and 1994 showed that the benthos has not recovered even after a decade. The investigated macroepibenthic community, which con­sists largely of filter and suspension feeders, plays an important role in stabilizing the entire ecosystem. Due to re­peated disturbances the function of this community has been impaired. The reaction of the benthos to these distur­bances shows that such communities serve not only to accurately interpret current conditions, hut also serve as a memory of past events. More effort should be made to decipher the rich information that the benthos provides. Keywords: benthos, macrofauna, North Adriatic Sea, Gulf of Trieste, disturbance, mortality, recolonization, anoxia, marine snow, long-term Ključne besede: bentos, makrofavna, Severni Jadran, Tržaški zaliv, motnje, smrtnost, rekoionizacija, anoksija, morski sneg, dolgoročne spremembe INTRODUCTION The Northern Adriatic Sea can be classified as a sen­sitive ecosystem. As the northernmost pari: of the Medi­terranean, it is subject to high annual fluctuations of phy­sical parameters such as temperature and salinity. Its shallow depth (< 30 m), strong stratification, high rive­rine input, along with pressure from tourism and com­mercial fishing contribute to making it a classical ex­ample of an endangered marine ecosystem. In addition to the above features, the Northern Adriatic is also sus­ceptible to other phenomena such as oxygen deficien­cies and marine snow production, which have taken on dramatic proportions in recent years (Brambati, 1988; Stachowitsch et ai., 1990; Volienweider & Ri na Idi, 1995). Despite the above sources of instability, the sublitto­ral soft bottoms of the Northern Adriatic are character­ized by well-developed benthic communities. The ben­thic communities in this area have been studied for al­most a century and have been summarized in regular intervals (Vatova, 1949; Péres, 1967; Camulin-Brida, 1974; Ott, 1991). These communities consist not only of a wefi-devefoped infauna, but also of a characteristic macroepifauna (Orel & Mennea, 1969). One of the most widespread epibenthic communities is the O-R-M community, named after three dominant genera, the brittle star Ophiothrix, the sponge Reniera, and the as­cidian Microcosmus (Fedra et al., 1976). This commu­nity consists largely of mobile and sessile filter or sus­pension feeders which are aggregated in the form of so-called multi-species clumps. During the course of ecological investigations in the Michae l STAC H O WITSCH , Alexande r FUCHS : ION C TER M CHA N OES I N TH E BENTHO S O F TH E NORTHER N ADRIATI C SEA , 7-1 6 1970s, this community was found to maintain a stable biomass and structure. The community plays an impor­tant role in stabilizing the entire ecosystem by removing suspended material from the water column and storing it in the form of benthic biomass (Ott & Fedra, 1977; Ott, 1981). However, a series of recent disturbances includ­ing oxygen deficiencies and massive marine snow de­velopment have overwhelmed this stabilizing function and led to the collapse of the benthos {Stachowitsch, 1984). These developments have been increasingly re­lated to eutrophication (Rosenberg, 1985; UNESCO, 1988). While such perturbations may be visible in the pelagic system in the form of reduced transparency, dis­coloration of the water, surface layers of mucus, and other short-term clues, the impact on the benthos is much more long-lived. Long-term studies are necessary to accurately detect the impact of such disturbances on the benthos and to determine how various short-term perturbations influence the overall system. MATERIAL AND METHODS In 1974 and 1975 the benthic communities in the Gulf of Trieste / Northern Adriatic Sea were investigated by means of an underwater TV-camera sled consisting of a 1" Vidicon black and white videocamera, a Hasselblad 500-el camera and a series of lamps (for a detailed de­scription of the system, see Machan & Fedra, 1975). Twelve profiles with a total length of 80 km and cover­ing an area of about 200 km2 were examined in order to define the borders of the O-R-M community (Fedra et ai., 1976). Additional profiles were also made in the Italian sector of the gulf to define adjoining communities (Fedra, 1978). This, coupled with a series of samples taken between 1973 and 1977 at a central position in the O-R-M community (station 1, Fig.1 in Stachowitsch, 1984) provided information on the undisturbed (premortality) condition of the benthic community. in 1983 a mass mortality event was observed at twelve stations in the Gulf of Trieste from 12-2 6 Sep­tember. One station (station 1) was revisited on six sepa­rate days in order to document the course of mass mor­tality. Photographs were taken with a NIKONOS II and NIKONOS IV-A camera equipped with a specially de­signed electronic flash and a NIKON SB-101 flash, re­spectively. ILFORD PAN F and KODACHROME 64 film was used. Samples for species identification were col­lected by hand. The recolonization process after the 1983 mass mortality event was investigated by taking 4 to 10 1 m2 macroepifaunal samples each year using SCUBA. All or­ganisms or biogenic structures on or projecting from the sediment were collected by hand and placed into 1 mm2 mesh bags. An effort was also made to collect the fauna lying immediately beneath the sediment surface. The samples were fixed in a 4% formafdehyde:seawater so­lution. The recolonization process described in this contribution is based on a single, randomly chosen 1 m2 sample for each year between 1984 and 1994. Wet weights of the macrofauna (> 1mm) were determined with a Sartorius H 120 electronic balance (+/- 0.001 g ). Wet weight measurements include mollusc shells but omit serpulid tube weights. RESULTS The O-R-M community: The large-scale underwater TV-camera sled survey showed that the O-R-M community covered a large area of the Gulf of Trieste, Based on the observations of Czi­hak (1959) and Riedl (1961), who reported high densi­ties of the brittle star Ophiothrix quinquemaculata off Rovinj, Croatia, this community is thought to extend far down into the Northern Adriatic. The evaluation of the videofilms showed that the designating genera Ophio­thrix, Reniera, and Microcosmus were not only biomass dominants, but also visually characterized the commu­nity. These genera, along with a wide range of other sponges and ascidians, were clearly aggregated in the form of multi-species clumps (Fedra ef a/,, 1976), with the intervening sediment surface containing scattered deposit feeders and carnivores. These multi-species clumps intitially grow on a biogenic base or "nucleus" consisting of gastropod shells (Murex brandaris, Truncu­lariopsis trunculus, Aporrhais pes-peiecani), bivalve shells (Area noae, Chlamys spp.), or sea urchin tests (Schizaster canaliferus, Psamrnechinus microtubercula­tus). The aggregated biomass in the O-R-M community shows that fixo-sessile, hemi-sessiie, and mobile species require such substrates. In the intact community, the former include sponges and ascidians and constitute the underlying structure of the multi-species clumps. Typical hemi-sessile species are Cucumaria planci and Chlamys varia, while the latter group includes the most visible mobile species, O. quinquemaculata (Fig. 1). SCUBA-diver-samples taken before, during, and after the TV-sled profiles showed the O-R-M community to be characterized by a high macroepifauna biomass of 370 g wet weight/m2, with maxima of more than 1000 g wet weighi/m2. The designating group, Ophiothrix- Reniera-Microcosmus, comprised 64% of the total biomass, with O. quinquemaculata alone contributing 28%. In con­trast, the mean biomass outside the community borders (the "peripheral areas" of Fedra et al., 1976) was evalu­ated at 166 g wet weight/m2, with O. quinquemaculata contributing only 6% to the total biomass. The benthos outside the O-R-M community borders was character­ized by a distinctly lower biomass, different species composition, smaller multi-species clumps, and a pre­dominance of mobile deposit feeders (hermit crabs, holothurians, and sea urchins). For a more detailed de­ Michae l STACHOWITSCH , Alexande r FUCHS : ION C TER M CH / i N TH E BENTHO S O F TH E NORTHER N ADRIATI C SEA , 7-1 & scription of other communities in the shallower parts of the Gulf of Trieste, see Fedra {1978). Mass mortalities A series of mass mortalities was observed in the course of our investigations on the benthos of the Northern Adriatic Sea over the past 20 years. The first mortality was documented in September 1974 during the TV-camera sled profile work. This area of decaying organisms was termed the "graveyard phenomenon" and affected the central region of the Gulf of Trieste (Positions 28 and 36, Fig. 1 in Fedra et a/., 1976). The exact boundaries of the affected area could not be de­termined at that time because they extended beyond the territorial waters of former Yugoslavia. However, the composition of the decaying organisms clearly showed that this area was occupied by the O-R-M community. in 1983 a renewed mass mortality event was ob­served at the central station in the gulf (station 1 = posi­tion 39 in Fedra etai, 1976). At this location, the entire course of ecosystem collapse, from behavioral modifi­cations to the sequence of mortality was documented and described in detail by Stachowitsch (1984). The af Fig. 1: Large, intact multi-species clump consisting of horny sponge, colonial ascidian and Cucumaria planci (top, middle), serpulid tubeworms and hydrozoans (top, right), and hermit crab in Trunculariopsis trunculus shell (center). Sponge serves as a substrate for a dense aggregation of the brittle star Ophiothrix quinquemacu­lata. Gulf of Trieste, 25 m. Slika 1: Velika, v naravnih razmerah živeča večvrstna skupina, ki jo sestavljajo roži čast a spužva, kohnijski kozolnjak, brizgača Cucumaria planci (zgoraj, v sredi­ni), serpuiidni mnogoščetinci in trdoživnjaki (zgoraj, desno) ter rak samotar v hišici čokatega voleka Truncu­lariopsis trunculus (v sredini). Spužva deluje kot pod­laga za čvrsto spojitev pegastega kačjerepa vrste Op­hiothrix quinquemaculata. Tržaški zaliv, 25 m. Fig. 2; Moribund Squilla mantis on sediment surface in Sept 1983. In background, decaying, mucus-covered multi-species clump. Slika 2: Morska bogomoljka Squilla mantis na površju morskega dna septembra 1983. V ozadju s sluzom prekrita razpadajoča večvrstna skupina. fected area was estimated to measure 250 km2 and cov­ered most of the bottom below 18 m in the gulf. Both the infauna and epifauna were affected by the 1983 mortality event. All organisms showed modified be­havior before death occurred. Among the infauna, the most common visible reaction was emergence from the sediment. At a later stage, most of these species were ob­served motionless, in a moribund state, on the sediment (Fig. 2), although some left short tracks on the surface (the heart urchin Schizaster canaliferus, the gastropod Aporrhais pes-pelecani) or were recorded swimming in the water column (Squilla mantis). Other reactions inclu­ded evisceration (Thyone fuscus), aggregating on mounds (Amphiura chiajei'), as well as gaping valves and ex­tended syphons in bivalves. Furthermore there was a di­stinct sequence of emergence and death in the in-fauna, with burrowing crustaceans (Upogebia tipica, jaxea nocturna) and echinoderms (S. canaliferus) emerging first, followed by polychaetes, sipunculans (Colfingia elongata, Sipunculus nudus), gastropods, and bivalves. The initial behavioral modifications of most of the mobile epifauna involved aggregating on elevated sedi­ment structures (mounds of burrowing crustaceans). Specific, characteristic behaviors of individual species included eviscerated hoiothurians (Holothuria tubulosa), the humped posture of the brittle star Ophiura texturata, overturning sea stars (Astropecten sp.) and, in the case of hermit crabs, the abandonment of the occupied shells by the crabs along with the detachment of symbiotic anemones (Calliactis parasitica). A key development was the destruction of the char­acteristic structure of the O-R-M community. Specifi­ Miclwe i STACHOWSTSCH , Alexande r FUCHS : l.ON C TER M CHANGE S I N TH E BENTHO S O F TH E NORTHER N ADRIATI C SEA , 7-1 6 Fig. 3: Close-up of decaying multi-species clump in Sept.1983. Adult O. quinquemaculata have left the clump> while dead, juvenile brittle stars are still at­tached to decomposing sponge substrate (top, left). As­sociated fauna includes the bivalve Chlamys varia cov­ered with decaying, encrusting sponge (center) and tu­beworm (left). Slika 3: Bližnji posnetek razpadajoče večvrstne skupine septembra 1983. Odrasli pegasti kačjerepi vrste O. quinquemaculata so že zapustili skupino, medtem ko so poginuli mladostni osebki Še vedno pritrjeni k raz­padajoči spužvi (zgoraj, levo). Združena favno sestav­ljata pokrovača vrste Chlamys varia, prekrita z razpa­dajočo spužvo (v sredini), in mnogoščetinec (levo). cally, the multi-species clumps were very susceptible to the mortality phenomenon. Sponges, for example, which were a major component in many multi-species clumps, were among the first species to die and become discoloured (Fig. 3). This had an immediate effect on the other species in the clumps, especially those species that use sponges as a substrate or habitat. Thus, motile species such as the designating species Ophiothrix quinquemaculata left the multi-species clumps on which they were normally aggregated and lay overturned on the sediment. This mortality event was coupled with large amounts of marine snow in the water column. As relatively large structures projecting from the sediment, the sponges in multi-species clumps snagged this drifting mucus material. The decaying sponges and marine snow caused the small fauna (e.g., nestlers and sponge dwell­ers) to emerge. For example, a wide range of dead crus­taceans such as Pisidia longicornis, Pilumntis spinifer, small shrimp, and amphipods were observed entangled in mucus-covered sponges. After the onset of the oxygen deficiency-induced be­havioral modifications, the collapse of the community proceeded very rapidly. Thus, the time interval between first unusual behavior and the onset of mortalities in the individual species was very rapid. Within 4 days over 90% of the macroepifaunal biomass was destroyed and most of the emerged infauna species began to die. Fourteen days later, at the end of the investigation pe­riod (26 Sept. 1983), only few living organisms were re­corded and the sediment was covered with a dark layer of decaying organic material. RecoJonization Compared with the rapidity of the collapse of the O ­R-M community in 1983, recolonization is a long-term process. Our investigations over a period of more than 10 years show that the community still has not recov­ered from this initial disturbance. This conclusion is based on total macroepibenthic biomass and the percent contribution of the former designating species (Fig. 4), A key to the reestablishment of community structure is the growth of the formerly characteristic multi-species clumps. A prerequisite for this is the availability of biogenic structures on which the larvae of sessile and motile epifauna species can settle. Not all substrates are suitable for epigrowth, and many factors can render po­tentially suitable ones unsuitable. The most conspicuous structures on the sediment surface after the 1983 and 1988 mortality events were bivalve shells and sea urchin tests. New epigrowth on such structures is heavily influ­enced by sedimentation. Sedimenting particles can rapidly cover smaller substrates (Fig. 5) and the top sur­face of even larger substrates such as the shells of the scallop Pecten jacobaeus can be covered by sealing particles. In such cases, the renewed growth of sessile organisms typically begins along the edges or from the underside. The first species registered after the 1983 mass mor­tality event differed from the structurally dominant forms previously recorded in the intact O-R-M community. The initial settlers on dead bivalve shells or sea urchin tests were encrusting bryozoans (e.g., Schizoporella sp., Schizomavella sp.) and serpulid tubeworms such as Po­matoceros triqueter and Serpula vermicularis. The tubes of the serpulfds eventually extended away from the sub­strates in an upright position, enlarging the "clumps'1 and giving them a more three-dimensional structure. These erect tubes favored the settlement of other epigrowth. Many juvenile bivalves {e.g., Hiatella arctica), as well as rapidly growing ascidians (Ciona intestinalis) and bydro­zoans were found between the serpulid tubes. In later successiona! stages, slower-growing ascidians (Micro­cosmus spp.) successfully settled. This led to a corres­ponding increase in the species associated with the thick tunic of these more long-lived ascidians (e.g., the bivalve Musculus subpictus). These developing multi-species clumps consisted not only of epifauna, but also of many mobile and nestling species like polychaetes, crustaceans, sipunculans and nemertines. These stages of Michae l STACHOWfTSCH , Alexande r FUCH5 : LON G TER M CHANGE S I N TH E BENTHO S O f TH E NORTHER N ADRIATI C SEA , 7-16 Fig. 4: Recolonization of benthic community between 1984 and 1994, based on 1 m2-sample taken in September of each year. Total macroepifaunal biomass and contribution of designating genera (stippled: Ophiothrix, Reniera, Microcosmus). The community has been subject to repeated disturbance and has not recovered even after 10 years. Slika 4: Rekolonizacija benloške združbe med letoma 1984 in 1994 na osnovi vzorca (s površino 1 m2) v sep­tembru vsakega leta. Skupna biomasa makroepifavne in prispevek posameznih rodov (pikčasto: Ophiothrix, Re­niera, Microcosmus). Združba je bila izpostavljena nenehnemu vznemirjanju in si ni opomogla celo po desetih letih. community recolonization clearly represent a succe­ssion. A detailed distinction and description of the indi­vidual stages of succession is the focus of ongoing research. Renewed disturbances The normal succession process was hampered sev­eral times in the course of the decade. The first distur­bance of the recolonization of the O-R-M community is related to the scallop Pecten jacobaeus which, along with the pen shell Pinna sp., established themselves in greater numbers on the sediment surface. Both species may represent distinct stages in the recolonization of the bottom. Pinna, for example, was rarely recorded in the samples of the premortaiity community. By 1987, the scallop population reached a size large enough to sup- Fig. 5: Layer of sediment partially covering fresh Laevi-port commercial fishing operations. The subsequent use cardium oblongum shell. The sea urchin Psammechinus of bottom trawls scarred the sediment and even crushed microtubercuiatus grazing on right valve. New epig-the scallops themselves {Fig. 6). More important with re­rowth typically begins on underside or edges of such gard to the reesîabJishment of the O-R-M community, fresh substrates. these operations overturned or broke apart many of the newly established multi-species clumps. We consider Slika 5: Plast usedlin delno prekriva še svežo lupino this activity to have played a roie in the biomass de­ srčanke vrste Laevicardum oblongum. Mali morski je­ žek Psammechinus microtubercuiatus na paši. Nova crease to 61 g wet weighi/m2 in 1987 (Fig. 4). One of the most visible effects of bottom trawling was on the obrast se značilno začenja na spodnji strani ali robovih še svežih podlag te vrste. Pinna population. The opening of younger Pinna lies Michael 5TACHOWIT5CH . Alexander FUCHS : LON G TER M CHANGE S I N TH E BENTHO S O F TH L NORTHER N ADRIATI C 5EA, 7-16 Fig. 6: Pectén jacobaeus shell broken apart and dis­lodged by bottom trawling gear. Initial growth of ser­pulids on top valve, arms of Ophiothrix quinquemacu­lata projecting from underside of lower valve. Two anemones (Calliactis parasitica) between the valves are probably associated with a hermit crab. Slika 6: Lupina velike pokrovače Pecten jacobaeus, ki jo je iz dna iztrgala in razbila globinska vlečna mreža. Začetna rast serpulidnih mnogoščetincev na gornji lu­pini, kraki kačjerepa vrste Ophiothrix qiuquemaculata štrlijo s spodnje strani spodnje lupine. Dve stražni vetrnici med (Calliactis parasitica) lupinama sta najbrž v združbi z rakom samotarjem. flush with the sediment surface and the bivalves are in­conspicuous. Later, the valves project above the sedi­ment, with the outer surfaces serving as a further sub­strate for epigrowth. The fishing gear broke off the pro­jecting part of the shells, killing the bivalves (Fig. 7). These empty shells subsequently served as a substrate for a dense epigrowth consisting largely of serpulid tu­beworms (both on the inner and outer valve surfaces). The effect of ongoing, regular trawling activity was clearly evident: the gear continued to shear off the ends of the valves and often snagged on the tubeworms ex­tending from the apertures. The force is sufficient to pull the deeply buried bivalves from the sediment (Fig. 8). A renewed, small-scale mortality, was recorded at the long-term station in September 1988. Based on diver-taken notes, underwater photography, and samples taken at adjoining stations, the area of this mortality was estimated to be 4 km2 . The seafioor was characterized by "black spots" indicating the position of small multi-species clumps and decaying organisms (Fig. 9). The bottom was also littered with fresh sea urchin tests (Psammechinus microtuberculatus, S. canalíferas) and numerous fresh bivalve shells (Laevicardium oblongum, Cardium sp.). The biomass in the September sample was reduced to 2 g wet weight/m2 and consisted of 3 Epi­ - Fig. 7: Freshly killed Pinna sp. with decaying soft parts still within the shell. Top part of valves sheared off by bottom trawl (Sept. 1987). Slika 7: Pred kratkim uničen leščur Pinna $p., z razpa­dajočimi mehkimi deli še vedno v lupini. Gornji del lu­pine je očitno odrezala globinska vlečna mreža (sept. 1987). zoanthus arenaceus polyps, 1 sipunculan and 10 poly­chaetes. This small-scale event, which occurred within the area affected in 1983, most likely did not impede the overall recolonization process in the North Adriatic, b«t did directly affect our long-term sampling station in the Gulf of Trieste. In 1988, 1989, and 1991, marine snow ("mare spor­co") events were registered in the Adriatic. The 1989 event had a major impact on the entire Northern Adriatic Sea. Marine snow settled on the bottom in the form of macroflocs, stringers, and clouds (Stachowitsch ef a!., 1990) as early as June 14. The amount of sinking material was sufficient to cover multi-species clumps (Fig. 7 in Stachowitsch ef a/., 1990) and to interrupt the suspension-feeding posture of entangled O. quin­quemaculata, for example. Interestingly, no severe an­oxia or widespread mortality occurred at the sampling station itself. It should be noted, however, that the benthos in 1989 was already highly reduced (18 g wet weight/m2) due to the mortality of the previous year. At the same time, the wide-ranging creamy and gelatinous surface layers (Stachowitsch et al., 1990) that covered large parts of the Northern Adriatic in summer 1989 ap­parently settled to the bottom further to the south, where they were associated with extensive mass mortalities over an area of 3000 to 4000 km2 in winter 1989/90 (Fig. 13 in Ott, 1991). Between 1990 and 1994 no major disturbances other than the ongoing bottom trawling operations were re­corded. Nonetheless, the recolonization process did not lead to the re-establishment of the O-R-M community. Michae l STA C H O W1TSCH , Alexander FUCHS : i.ON C TER M CHANGE S I N TH E BENTHO S O F TH E NORTHER N ADRIATI C SEA , 7-1 6 Fig. 8: Pinna sp. dislodged from the sediment by ben­thic trawl fishing equipment snags on projecting part of shell along with its epigrowth. Crushed serpulid tube-worms and Chlamys varia lying on the sediment sur­face. Note presence of predators Trunculariopsis trun­cal us and Psammechinus m icro luber cu I a tus, right) feeding on detached epigrowth. Note also C. varia attached to freshly exposed part of Pinna sp. Slika 8: ieščur Pinna sp., ki ga je iz podlage za štrleči del skupaj z obrastjo izvlekla globinska vlečna mreža. Na površju morskega dna ležijo uničeni serpulidni mnogoščetinci in mala pokrovača Chlamys varia. Z lo­čeno obrastjo se hranita plenilca - čokati volek Truncu­lariopsis trunculus in mali morski ježek Psammechinus microtubercuiatus (desno). K sveže odtrganemu delu leščurja je pričvrščena mala pokrovača. Specifically, neither the biomass nor the designating species reached former leveis (fig. 4). The overall low biomass reflects the smaller size of multi-species clumps vs. pre-mortaltfy values.The maximum biomass of indi­vidual clumps measured between 1990 and 1994, for example, was 9.5 g, with average values below Ig . Multi-species clumps can reach large dimensions. Wur­zian (1977) classified these aggregations into weight classes of 0 to 80 g, 80 to 240g, 240 to 560g, and > 560g. No multi-species-ciump of the larger weight clas­ses were found in the ten years following the 1983 mor­tality. In fact, no 1 m2 sample even approached the wei­ght of a single, former medium- to large-sized clump. Although the percent contribution of the former designating group of species increased (for example from 0 to 39% between 1984 and 1987), the composi­tion within the group remained atypical. While Micro­cosmus began to reestablish itself, the sponge Reniera was virtually absent (present for example in the 1993 sample: 6.8 g) and only isolated specimens of Ophio­thrix quinquemaculata (mostly juveniles) were collected in the samples. Not one adult O. quinquemaculata, Fig. 9: The sea star Astropecten aurantiacus surrounded by several "black-spots" representing decaying organ­isms in Sept. 1988. Fresh test of the sea urchin Psam­mechinus microtubercuiatus (top, middle) and emerged gastropods (Aporrhais pes-pelecani, center). Slika 9: Oranžno morsko zvezdo Astropecten auranti­acus obkrožajo številni razpadajoči organizmi v obliki črnih pik (svpt. 1988). Sveže testiranje malega mor­skega ježka Psammechinus microtubercuiatus (zgoraj, v sredini) in pojavljajočih se polžev pelikanovih stopale (Apporrhais pes-pelecani, v sredini). which was formerly present in numbers of up to 500 individuais/m2 , was found in the samples evaluated between 1991 and 1994. DISCUSSION Due to the narrower range of fluctuations of physical parameters in marine versus terrestrial ecosystems, large-scale natural disturbances in the marine environ­ment are relatively uncommon and poorly documented. This is particularly true in sublittoral soft bottoms. When disturbances do occur, they tend to be most overtly manifested in the pelagic subsystem (e.g., discoloration clue to plankton blooms). In cases where disturbances in the pelagic subsystem are coupled with or trigger per­turbations in the benthic subsystem, the discrepancy between the duration of the phenomena in the two realms can be very large: the transient nature of pelagic disturbances stands in contrast with the long-term dis­turbances in benthic communities, especially in the case of a well-developed macrofauna. This makes benthic communities ideally suited to detect the actual effect of perturbations on the overall ecosystem. In the investigated O-R-M community, for example, the perturbations that triggered mass mortality, i.e., ma­rine snow events and anoxias in the water column, were short-term events on the scale of days, weeks, and Michael STACHOW1TSCH . Alexander FUCHS : LON G TERM CHANGE S IN TH E BENTHO S O f TH E NORTHER N ADRIATI C SEA, 7-16 months. The recolonization process {notwithstanding the renewed disturbances), on the other hand, clearly in­volves periods of years or even decades (Fig. 4). Ibi s period is on the longer end of the range reported for other sublittoral bottoms. Total recovery in the New York Bight after anoxia was expected to take several years due to the extensive area affected (Steimle & Ra­dosh, 1979). Recolonization in the Bornhoim Basin in the Baltic after a 1968 mortality led, even after several years, lo a community showing very little resemblance to the original Macoma baitica community (Leppakoski, 1971). Repeated disturbances complicate the process, leading to steady impoverishment (German Bight; Ra­chor, 1980) or the annual reestablishment of transitory successional stages (Chesapeake Bay; Officer ef a/., 1984). Compared with the communities mentioned above, which are dominated by infauna species, recovery in the O-R-M community is no doubt compounded by the complex structure and associations of the macro­epibenthic invertebrates (e.g., multi-species clumps). Fully developed multi-species clumps are perennial structures. A wide range of physical and biological conditions are necessary for mature aggregations to develop. Clearly, this proceeds via a number of succes­sional stages. Different types of multi-species clumps (e.g., sponge- versus ascidian-dominated) can arise de­pending on larval settlement factors, space competition, predation, etc. Their final appearance can further be modified by ecological interactions. For example, the distribution of O. quinquemaculata can be interrupted or modified when the sponges are occupied by the crab Pilumnus spinifer; in this case, crab number and clump size determine the small-scale distribution pattern of the brittle star (Wurzian, 1977). A second path of multi-species-clump formation, in­volving hermit-crab occupied gastropod shells, is also known. Virtually ail gastropod shells go through the hermit crab population. Because they are transported over long distances by the crabs (Stachowitsch, 1979), these structures are sediment-and predator-free, main­tained in a stable orientation, and elevated above the bottom. These shells are therefore optimal substrates with manifold advantages for the epifauna. They serve as a substrate for a diverse community consisting of more than 120 species and a wide range of functional groups and are essentially mobile multi-species clumps (Fig. 10). When these shells are eventually deposited by the crab - after they have been modified by a complex interplay of "constructive" epibionts and "destructive" endolithic forms - the complex community of "sym­bionts" can survive to form typical multi-species clumps {Stachowitsch, 1980). Fully developed multi-species clumps in the intact community are important functional units. 87.5% of the biomass consists of filter-and suspension-feeding organi- Fig. 10: "Mobile multi-species dump" consisting of large epigrowth carried by the hermit crab Paguristes eremita (eyes and antenna of crab visible, bottom cen­ter). Epigrowth consists of two ascidians (Microcosmus vulgaris) and the bivalve Area noae, the latter bearing several small ascidians (Distomus variolosus). Ascidians serve as substrate for brittle stars. Slika 10: "Gibljiva" večvrstna skupina, sestoječa iz veli­ke obrasli, ki jo nosi rak samotar Paguristes eremite (lepo vidne so njegove oči in tipalnice, spodaj v sre­dini). Obrast sestoji iz dveh kozolnjakov (Microcosmus vulgaris) in noetove barčice Area noae, na kateri je več majhnih kozolnjakov vrste Distomus variolosus. Ko­zolnjaki dajejo podlago kačjerepom. sms. In such shallow environments, the benthos is not merely a receiving compartment. Rather, complex feed­back processes, with the benthos controlling the pela­ges, are in effect (Figs 11, 12 in Ott, 1991). Ott & Fedra (1977) estimated that the suspension feeders in the Gulf of Trieste can remove all the suspended material in the water column every 20 days. This is on the same order of magnitude as calculated for the Oosterschelde (Herman & Schoiten, 1990), Swedish waters (Loo & Ro­senberg, 1989), the USA {Cloern, 1982), and France (Hily, 1991). Such communities have therefore been termed a "natural eutrophication control" {Officer ef a!., 1982). They play a key role in the stability of the entire ecosystem, and their loss makes the system more sensi­tive to perturbations. The current status of the O-R-M community, with its low biomass, atypical species com­position, and altered structure, makes it unlikely that it fulfills its premortality regulatory capacity. This stabilizing function is hampered by the above-mentioned marine snow and oxygen deficiency events. This is compounded by the continuous threat to com­munity structure due to bottom trawling. Such fishing operations are known to have a negative impact on benthic communities {de Groot, 1984; Langton & Michae l STA C H O WITSCH , Alexande r FUCHS : LON G TER M CHANGE S I N TH E BENTHO S O r TH E NORTHER N ADRIATI C SOA, 7-1 5 Robinson, 1990). This activity is particularly destructive in the Guif of Trieste, where fishermen can use land­marks to very efficiently space their trawls. The effect is clearly visible on the bottom due to furrow-like mark­ings, destroyed fauna {Figs 6, 7), as well as the repeated demolition of underwater experiments. It is also respon­sible for unnatural features such as uprooted Pinna shells. Only such fishing gear, which snags on the ser­pulid tubeworms extending from the bivalve's aperture, is capable of extracting these deeply buried bivalves from the sediment (Fig. 8). Clearly, detailed knowledge of benthic communities, preferably including direct observations, can provide valuable information on the status of the overall system. Three criteria or levels can be called upon to deliver such information: 1) behavior, 2) mortalities, and 3) recolonization. 1.) On the first level, behavioral criteria provide in­formation on an ongoing disturbance. The time scale in­volves hours, days or weeks. Here, the severity and du­ration of oxygen deficiency can be gauged by the spec­trum of species exhibiting unusual behavior and by the type of behavior exhibited by these species. 2.) On the second level the course of mortalities or the more immediate post-disturbance community com­position of surviving species serve as a criteria to gauge anoxias. The time scale is weeks, months, or up to a year or more. The post-mortality absence of only sensi­tive species or age classes would indicate less severe oxygen depletions, while mortalities of more resistant forms would indicate more severe conditions. 3.) On a third, more long-term level, recolonization processes can provide information on earlier, perhaps undocumented oxygen crisis. The time scale here may involve years, and as in the case of the Northern Adriatic, even decades. A prerequisite for optimal re­constructions is knowledge of former community com­position, of opportunistic species, and of successions! and possible climax stages. Ideally, knowledge of these processes would allow the time of ecosystem collapses to be pinpointed with accuracy even years after the event. Thus, benthic communities can serve as the memory of recent disturbances (Stachowitsch, 1992). Ultimately, benthic communities also provide informa­tion on a fourth level, i.e., the paleontological/ geological record as well. Many shallow-water marine ecosystems have com­mon physical features and are subject to a similar range of threats and stresses; the reactions of macrobenthos also show parallel features worldwide (Pearson & Ro­senberg, 1978; Stachowitsch & Avcîn, 1988). This is im­portant in recognizing and gauging such threats and suggesting solutions. The events in the Northern Adriatic Sea are therefore of high predictive value for endan­gered coastal waters worldwide. More effort should be made to decipher and utilize the information that ben­thic communities provide. ACKNOWLEDGEMENTS This study was supported by project Nr. 67138 of the Austrian "Fonds zur I-'orderung der wissenschaftlichen Forschung" to Prof. Rupert Riedl and M.S. and by a se­ries of stipends from the University of Vienna (Inter­national Office) to M.S. and A.F. The camera equipment was financed by a grant from the "Hochschuljubilaums­stiftung der Stadt Wien" to M.S. The authors gratefully acknowledge the long-term hospitality extended and facilities provided by the director and staff of the Marine Biological Station Piran as well as the continued support of Prof. Jôrg Ott, University of Vienna. POVZETEK Severni del Jadranskega morja je tako kot mnoge plitve obrežne vode izpostavljen številnim obremenitvam in nevarnostim, med katere sodijo pomanjkanje kisika, morski sneg in ribištvo. Zaradi množičnega umiranja organizmov leta 1983 v tem delu Jadrana je propadla bentoška združba z močno razširjenim arealom. Dolgoročno zbiranje vzorčnih primerkov med letoma 1984 in 1994 je pokazalo, da si tamkajšnji bentos ni opomogel celo v obdobju desetih let. Raziskovana makroepibentoška združba, sestoječa predvsem iz organizmov, ki se hranijo s filtratom, igra pomembno vlogo v stabilizaciji celotnega ekosistema. Nenehne obremenitve v severnem Jadranu so močno oslabile delovanje te združbe. Reakcije bentosa na motnje kažejo, da takšne združbe niso le natančen kazalec obstoječih razmer, marveč tudi nekakšno skladišče informacij o preteklih dogajanjih. Avtorja menita, da bi se bilo zatorej treba bolj potruditi pri razvozlavanju informacij, ki nam jih daje severnojadranski bentos. REFERENCES Trieste-Roma; 68 pp. Cloern, J.E. 1982. Does benthos control phytoplankton Brambati, A. (Ed.), 1988. il fenomeno del "mare sporco" biomass in south San Francisco Bay? Mar. Ecol. Prog, neil'Adriatico. Consiglio Nazionale delle Ricerche, Pro-Ser. 9:191-202. gretto Strategico Oceanografia e Tecnologie Marine. Czihak, G. 1959. Vorkommen und Lebensweise der Michae l STACHOWITSCH , Alexande r FUCHS : LON G TER M CHANCE S I N TH E 8SNTHÖ S O F TH E NORTHER N ADRIATI C SEA , 7-1 6 Ophiothrix quinquemaculata in der nördlichen Adria bei Rovinj. Thalassia Jugosi., 1(7):19-27. de Groot, S.J. 1984. The impact of bottom trawling on benthic fauna of the North Sea. Ocean Management 9:177-190. Fedra, K. 1978. On the ecology of the North Adriatic Sea. Wide range investigations on the benthos: The Gulf of Trieste. Memorie di Biogeogr. adriat,Suppl.(!X): 69­ 87. Fedra, K., Ölscher, E.M., Scherübel, C., Stachowitsch, M. & Wurzian, R.S. 1976. On the ecofogy of a North Adriatic benthic community: Distribution, standing crop, and composition of the macrobentbos. Mar. Biol., 28{2):129-145. Gamulin-Brida, H. 1974. Biocenoses benthiques de la Mer Adriatique. Acta adriat., 15(9):1-102. Herman, P.M.J. & Schölten, H. 1990. Can Suspension feeders stabilize estuarine ecosystems? In: M. Barnes & R.N. Gibson (eds.): Trophic relationships in the marine environment, pp. 104-116. Aberdeen Univ. Press, Aber­deen. Hily, C. 1991. Is the activity of benthic suspension feed­ers a factor controlling water quality in the Bay of Brest? Mar. Ecol. Prog. Ser. 69,179-188. Langton, R.W. & Robinson, W.E. 1990. Faunal associa­tions on scallop grounds in the western Gulf of Maine. J. Exp. Mar. Biol. Ecol. 144:157-171. Leppäkoski, E. 1971. Benthic recolonization of the Bornholm Basin (Southern Baltic) in 1969 - 1971. Tha­lassia Jugosi.,7:171-179. Loo, L.-O. & Rosenberg, R. 1989. Bivalve suspension- feeding dynamics and benthic-pelagic coupling in an eutrophicated marine bay. j. Exp. Mar. Biol. Ecol. 130:253-276. Machan, R. & Fedra, K. 1975. A new towed underwater camera system for wide-range benthic surveys. Mar. Biol. 33:75-84. Officer C.B., Smayda T.J., Mann, R. 1982. Benthic filter feeding: a natural eutrophication control. Mar. Ecol. Prog. Ser. 9:203-210. Officer C.B., Biggs R.B., Taft, J.L., Cronin, L.E., Tyler, M.A. & Boynton, W.R. 1984. Chesapeake Bay anoxia; origin, development, and significance. Science 223:22­27 Orel, G., Mennea, B. 1969. 1 popolamenti bentonici di alcuni tipi di fondo mobile del Goifo di Trieste. Pubbl. Staz. Zool. Napoli, 37(2 suppi.):261-276. Ott, J., Fedra K. 1977. Stabilizing properties of a high biomass benthic community in a fluctuating ecosystem. Helgoländer wiss. Meeres unters. 30:485-494. Ott, J. 1981. Adaptive strategies at the ecosystem level: examples from two benthic marine systems. P.S.Z.N.I: Mar. Ecol. 2:113-158. Ott, J. 1991. The Adriatic benthos: problems and per­spectives. In: G. Colombo, J. Ferrari, V.U. Ceccherelfi & R. Rossi (eds.): Marine eutrophication and population dynamics. 25th Europ. mar. Biol. Symp. Ferrara, pp. 77­ 84. Olsen & Olsen, Fredensborg. Pearson, T.H. & Rosenberg, R. 1978. Macrobenthic suc­cession in relation to organic enrichment and pollution of the marine environment. Oceanogr. mar. Biol. a. Rev. 16:229-311. Peres, J.M. 1967. The Mediterranean benthos. Ocean­ogr. mar. Biol. a. Rev. 5:449-533. Rachor, E. 1980. The inner German Bight - an ecologi­cally sensitive area as indicated by the bottom fauna. Helgolander wiss. Meeresunters. 33:522-530. Rtedl, R. 1961. Etudes des fonds vaseaux de J'Adriaiiq ue. Metbodes et resultats. Reel Trav. Stn mar. Endoume, 23:161-169. Rosenberg, R. 1985. Eutrophication - the future marine coastai nuisance? Mar. Poll. Bull. 16(6):227-231. Stachowitsch, M. 1979. Movement, activity pattern and role of a hermit crab population in a sublittorai epifauna community.). Exp. Mar. Biol. Ecol. 39:135-150 Stachowitsch, M. 1980. The epibiotic and endolithic species associated with the gastropod shells inhabited by the hermit crabs Paguristes ocuiatus and Pagurus cuanensis. P.S.Z.N.I: Mar. Ecol., 1(1):73-102. Stachowitsch, M. 1984. Mass mortality in the Gulf of Trieste: The course of community destruction. P.S.Z.N.I: Mar. Ecol., 5(3):243-264. Stachowitsch, M., & Avcin, A. 1988. Eutrophication-in­duced modifications of benthic communities, in: Eutro­phication in the Mediterranean Sea: receiving capacitiy and monitoring of long-term effects. UNESCO Rep. Mar. Sci. 49:67-80. Stachowitsch, M., Fanuko, N. & Richter, M. 1990. Mu­cus aggregates in the Adriatic Sea: an overview of stages and occurrences. P.S.Z.N.I: Mar. Ecol., 1 1:327-350, Stachowitsch, M, 1992. Benthic communities: eutrophi­cation's "memory mode". Science of the Total Environ­ment, Supp!.:101 7-1028. Steimle, F.W. & Radosh, D.J. 1979. Effects on the benthic invertebrate community. In: R.L. Swanson & C.J. Sindermann (eds.): Oxygen depletion and associated benthic mortalities in New York Bight, 1976, pp. 281­ 293. NOAA Professional Paper No. 11, pp. 281-293. UNESCO. 1988. Eutrophication in the Mediterranean Sea: receiving capacity and monitoring of long-term ef­fects. UNESCO reports in marine science 49, 195 pp. Vatova, A. 1949. i..a fauna bentonica dell'Alto e Medio Adriatico. Nova Thalassia f(3):T~110. R.A. Voilenweider & A. Rinaldi (eds.). 1995: Marine mucilages. The Science of the Total Environment, spe­cial issue, Vol. 165. Elsevier, Amsterdam. 235 pp. Wurzian, R.S. 1977. Predator - prey interaction between the crab PHumnus hirtellus (Leach) and the brittle star Ophiothrix quinquemaculata (D. Chiaje) on a mutual sponge substrate. In: B.F. Keegan, P.O'Ceidigh & P.J.S. Boaden (eds.): Biology of benthic organisms, pp. 613­ 620. Pergamon Press, Oxford.