Differences in the species- and size-composition of fish assemblages in three distinct seagrass habitats with differing plant and meadow structure

Fish faunas were sampled seasonally using a large and a small beam trawl in three seagrass habitats comprising predominantly Amphibolis giffithii or Posidonia sinuosa or Posidonia coriacea, which differ in seagrass and meadow structure. Amphibolis giffithii and P. sinuosa both produce a relatively dense leaf canopy, but the former exhibits a distinct architecture, with the leaf canopy overlying relatively open spaces surrounding woody stems, compared to the uniformly dense blade-like leaves of P. sinuosa which emerge directly from the sediment. In comparison, P. coriacea provides a landscape of patchy seagrass amongst areas of bare sand. Since the latter seagrass habitat contains large areas of sand, fish were also sampled in adjacent unvegetated areas. The number of species and density of fish were greater (P<0.05) in P. sinuosa than in either A. griffithii or P. coriacea. The mean number of species caught using the large trawl ranged from 16 to 24 in the first of these habitats compared to 14-21 and 9-15 in the last two habitats, respectively, and the mean densities ranged between 78 and 291 fish 1000 m(-2) in P. sinuosa compared to 31-59 fish 1000 m(-2) in A. griffithii and 31-59 fish 1000 m(-2) in P. coriacea. The biomass of fish was greater (P<0.05) in both P. sinuosa and A. griffithii than in P. coriacea (4.2-5.3 kg and 3.3-6.2 kg versus 0.7-1.9 kg 1000 m(-2), respectively). Furthermore, the size-structure of fish differed among these habitats, where the median weight of fish was 72.1 g in A. griffithii, compared to 7.5 g and 19.8 g in P. sinuosa and P. coriacea, respectively. Ordination and ANOSIM demonstrated that the species-composition differed markedly among the three seagrass habitats (P<0.001), suggesting that fish species display a distinct preference for particular seagrasses characterised by different architecture. Differences in species-composition among the seagrass habitats partly reflected the size-composition of fish in each habitat. The open space below the canopy of A. griffthii is likely to allow larger fish to occupy this habitat, whereas only small fish would be able to penetrate the dense foliage of P. sinuosa. Differences in species- and size-composition of fish among these habitats may be the result of settlement-sized larvae discriminating between particular seagrass and meadow structures, or fish being subject to different levels of predation and/or accessibility to food or space. The species-composition in P. coriacea was highly dispersed and did not differ from that of unvegetated areas. While several species were associated with both P. coriacea and bare-sand habitats, some species did display a high affinity with the seagrass P. coriacea. This may reflect an association with the sparse and narrower leaves of this seagrass or with the patchy occurrence of the seagrass Heterozostera tasmanica, which commonly occurs as an understorey in this habitat.