Patterns in the abundance and size-distribution of syngnathid fishes among habitats in a seagrass-dominated marine environment

Syngnathid fishes were sampled using a 1 m wide beam trawl during the day and night in each season from summer 1996/1997 to summer 1997/1998 in five habitat types in an approximately 90 km(2) area located on the lower west coast of Australia. The seagrasses Amphibolis griffithii, Posidonia sinuosa and Posidonia coriacea and shallow unvegetated sand were all in depths of 4-9 m, while deep habitat, which comprised mainly bare sand with isolated patches of the seagrasses Heterozostera tasmanica and Halophila ovalis, occurred at depths of 12-16 m. While A. griffithii and P. sinuosa each formed dense monospecific meadows, P. coriacea occurred in sparse clumps surrounded by areas of bare sand and patches of H. tasmanica. While catches of spotted pipefish Stigmatopora argus occurred mainly in P. sinuosa and P. coriacea, individuals of this species that exceeded ca. 55 mm in snout-vent length were far more abundant in the former habitat whereas smaller fish occurred mostly in the latter. Densities of S. argus were similar in P. sinuosa and P. coriacea, but differed between seasons, and a season/habitat interaction was present. In contrast, wide-bodied pipefish Stigmatopora nigra were collected mainly in P. coriacea and deep habitat and, although the densities were greater in P. coriacea than in deep habitat, the size-distributions of this species in these habitats were similar. Notably, S. nigra was never collected in P. sinuosa. Although less abundant than the above pipefish, the long-snouted pipefish Vanacampus poecilolaemus was collected almost exclusively in P. sinuosa. Few syngnathids were collected from shallow unvegetated sand or from A. griffithii, which differed markedly in plant structure from both Posidonia species. It is suggested that these syngnathid species occupy habitats that best enable them to remain inconspicuous to predators. Both S. argus and S. nigra have green or brown colouration and mimic strap-like seagrass leaves which they grasp with prehensile tails, whereas V. poecilolaemus, which is dark coloured and lacks a prehensile tail, most likely lies among the rich detrital material that accumulates beneath the dense P. sinuosa canopy. The movement of larger-sized S. argus from the narrow-leaved P. coriacea to the relatively broadleaved P. sinuosa may enable the adults of this species to be better camouflaged than if they had remained in the former habitat. As they grow to only approximately half the size of S. argus, S. nigra of all sizes could remain concealed in narrow-leaved seagrasses like P. coriacea and H. tasmanica. The presence of more S. nigra in P. coriacea compared with deep habitat probably reflected the greater expanse of seagrass canopy available in the former habitat, while the inability of either Stigmatopora species to mimic the short leaves of A. griffithii could help to explain their scarcity in this seagrass. The almost total absence of S. nigra and small S. argus from P. sinuosa could be due to their inability to grasp the relatively broad leaves of this seagrass or the presence in this habitat of predators that selectively predate these smaller pipefish. That very few small-sized S. argus were present in P. coriacea during winter 1997, despite the fact that males of both Stigmatopora species carried broods of young in all seasons, suggests that the reproductive output of this species and/or the survival of recruits varied during the study period. (C) 2003 Elsevier Science B.V. All rights reserved.