A. Schanz �� P. Polte �� H. Asmus Cascading effects of hydrodynamics on an epiphyte���grazer system in intertidal seagrass beds of the Wadden Sea Received: 8 October 2001 /Accepted: 18 February 2002 / Published online: 8 May 2002 �� Springer-Verlag 2002 Abstract This study examines experimentally how water movement may alter epiphyte���grazer systems in inter- tidal seagrass beds. Field observations in the Sylt���R��m�� Bay (German Wadden Sea, SE North Sea) showed that the biomass of seagrass epiphytes was highest on seag- rasses exposed to water movement, whereas at sheltered sites the epiphyte cover was negligible. In contrast, the seagrass shoot density and aboveground biomass was comparably sparse and the abundance of Hydrobia ulvae was extremely low at exposed areas, but showed maxi- mum values at sheltered seagrass beds. Cross trans- plantation experiments and enclosure experiments between sheltered and exposed seagrass beds showed that adhering snails were washed off from seagrasses soon after transplantation into an exposed seagrass bed, and epiphytes started to grow. After 4 weeks the epi- phyte biomass was similar to the that of the adjacent exposed seagrass bed. When heavily epiphytised seag- rasses were transplanted from exposed into sheltered areas, the epiphytes were completely grazed down by immigrating snails within a week. Experiments carried out by means of an in situ ������three-current-flume������, mod- ifying the entire current velocity, showed that snail density was significantly negatively correlated with in- creasing current velocity, whereas epiphyte biomass showed a significant positive correlation with current speed. These results suggest a cascading impact of hy- drodynamics on an epiphyte���grazer system in intertidal seagrass beds, by directly affecting the density of grazers and indirectly leading to enhanced epiphyte growth, thereby inhibiting seagrass development. Additionally it shows that cascading effects within the trophic web cannot only be triggered by biotic interdependencies, but can also be caused by physical factors. Introduction Epiphyte���grazer systems are characteristic elements of epibenthic communities, such as seaweed stocks and seagrass beds (Gambi et al. 1992 Bologna and Heck 1999). They play an important functional role, because a large part of the energy flow passing the community is dissipated through these systems and is thus transferred to higher trophic levels (Asmus and Asmus 1985, 2000 Klumpp et al. 1992). In some macrophytobenthic com- munities, especially in seagrass beds, epiphyte grazing tends to surpass or even outweigh the process of direct grazing of the higher plants (Klumpp et al. 1992), which often have developed mechanisms to prevent being grazed upon (Ravn et al. 1994 Vergeer and Develi 1997). Since epiphytes compete with seagrasses for light and nutrients (Sand-Jensen 1977 Shepherd et al. 1989), their regulation by grazers will initiate a cascading effect on the host plants, promoting the growth conditions of seag- rasses (e.g. van Montfrans et al. 1982, 1984 Orth and van Montfrans 1984 Howard and Short 1986 Orth 1992 Neckles et al. 1993 Jernakoff and Nielsen 1997 Valen- tine et al. 1997 Garcia et al. 1999), which is particularly of importance in eutrophicated areas (Williams and Ruckelshaus 1993 Philippart 1995). Despite extensive studies on factors acting directly or indirectly on grazing of epiphytes (Jernakoff et al. 1996), the effect of water movement on these complex animal���plant interactions has rarely been considered until now. Generally the roles of predation and habitat complexity as primary factors influencing abundance and distribution patterns of grazers have been regarded (Orth 1992), whereas the roles of physical factors have rarely been examined. However, previous studies have considered the direct impact of hydrodynamics on seagrasses (Gerard and Mann 1979 Fonseca and Kenworthy 1987 Koch 1994, 1999 Fons- eca and Bell 1998 Short and Neckles 1999 Robbins and Bell 2000 van Katwijk and Hermus 2000) and the in- fluences of mechanical factors on seagrass leaf fouling (Patriquin 1973 Fonseca et al. 1982 Jacobs et al. 1983). Marine Biology (2002) 141: 287���297 DOI 10.1007/s00227-002-0823-8 A. Schanz (&) �� P. Polte �� H. Asmus Alfred Wegener Institute for Polar and Marine Research, Wadden Sea Station Sylt, Hafenstrasse 43, 25992 List, Germany E-mail: aschanz@awi-bremerhaven.de Tel.: +49-4651-956102 Fax: +49-4651-956200

In the Wadden Sea, the mud snail Hydrobia ulvae is the dominant grazer on sheltered tidal flats and inter- tidal seagrass beds (Asmus and Asmus 1985, 1990 Dekker 1989 Asmus 1994 Reise 1994 Philippart 1995 Schanz et al. 2000). Mud snails are able to drift (Newell 1962 Armonies and Hartke 1995), and therefore a sen- sitivity of these organisms to hydrodynamics is expected. Based on field observations comparing a sheltered seagrass bed with low epiphyte coverage and a hydro- dynamically exposed Zostera bed heavily overgrown with epiphytes, it was hypothesised that the occurrence of epiphytes on seagrass leaves depends on the density of grazers, which in turn is controlled by the hydrodynamic regime. Field measurements were conducted to support these observations. To demonstrate experimentally how water movement may alter grazing processes in intertidal seagrass beds and may therefore play a fundamental role in seagrass bed development, cross-transplantation ex- periments with and without enclosures between sheltered and exposed seagrass beds were carried out. To quantify the impact of currents on the abundance of H. ulvae and their grazing e���ciency, an in situ flume, modifying the entire current velocity, was used. Materials and methods Study sites The studies were performed at two intertidal seagrass beds near the island of Sylt, which is situated in the German part of the Wadden Sea (North Sea) (Fig. 1). The investigated sites are part of a shal- low, nearly closed embayment between the mainland and the two islands Sylt and R��m�� (Sylt���R��m�� Bight), which are connected to the mainland by causeways. Predominant water circulation pat- terns are generated by a combination of both wind conditions and tides entering and leaving the bight through one comparatively narrow tidal inlet. Tides are semi-diurnal, with an average ampli- tude of 1.8 m. Salinity ranges from about 28 to 32 psu. Mean an- nual water temperature is 9��C, with a seasonal variation from 0��C to 19��C (see also Reise 1985 Bayerl and Higelke 1994 Reise and Lackschewitz 1998 Asmus and Asmus 2000 Asmus et al. 2000). Seagrass beds of the Sylt���R��m�� Bight occupy a relatively high share of the tidal flats (12% or 15.6 km2) (Asmus and Asmus 2000). Except for one small Zostera marina stand, they were pure Zostera noltii Hornem. stocks on sand and muddy sand during the inves- tigation period. Z. noltii beds emerge for 4���6 h per tidal cycle and grow under different hydrodynamic exposure conditions. Two Z. noltii beds (Fig. 1) were studied in July, August and September 1999 (the main vegetation period of the seagrass): an exposed Z. noltii bed (Fig. 1), where strong initial tidal currents from 0.20 up to 0.33 m s���1 were recorded during calm weather (during 1 h at the beginning of flood) entering the bay in an east���west direction, and which were probably due to the proximity of a deep tidal channel (Table 1) and a second, more extensive, sheltered Z. noltii bed (Fig. 1), which was protected by the island from waves, due to prevailing westerly winds. Here tidal currents from 0.06 m s���1 up to maximum values of 0.20 m s���1 were recorded (measured as men- tioned above) these currents were about half as strong as those measured at the exposed site (Table 1). Seagrass density and biomass To compare both seagrass sites, shoot density and aboveground biomass of Z. noltii were estimated in summer 1999. Shoot density was determined monthly in six replicate test areas using a frame (25��25 cm). Aboveground biomass of Z. noltii was calculated by taking box cores of 100 cm2 (n=6). Seagrass plants were clipped at Fig. 1. Location of the study area in the Wadden Sea near the island of Sylt, North Sea, Germany. Distribution of in- tertidal seagrass beds (shading) in the Sylt���R��m�� Bay. Arrows indicate the investigated hydro- dynamically exposed (E) and sheltered (S) seagrass sites (D: Germany DK: Denmark NL: The Netherlands) 288