Little is known about the relative influence of landscape structure on the spatial distribution and abundance of marine organisms. To address this problem, we applied landscape ecology concepts and methods, together with conventional sampling techniques and path analysis, to test alternative hypotheses of linkages between marine nekton and landscape structure in Moreton Bay, Queensland (Australia). We quantified substratum structure at 3 spatial scales: (1) whole landscape mosaic (10s of hectares); (2) habitat type (benthic class) (100s m 2 to hectares) and (3) within-patch scale (cm 2 to m 2 ). Substratum structure at all scales was important for assemblage density and number of species, with the landscape structure of individual habitat types explaining more of the spatial variation than either within-patch structure or the structure of the whole landscape mosaic. Density and the number of species of seagrass residents were strongly influenced by landscape composition quantified as the proportion of all seagrass habitat (r 2 = 0.40 and 0.48 respectively) and the proportion of long-leaved Zostera capricorni (r 2 = 0.34 and 0.30 respectively) seagrass in the landscape. An abrupt decline in assemblage density and number of species was evident at < 20% seagrass cover. More species of fish used mangroves with adjacent continuous seagrass beds than mangroves with adjacent patchy seagrasses or unvegetated sandflats. Several species of fish using mangroves at high tide were more strongly influenced by the spatial configuration of mangrove patches and the composition of adjacent substratum than the internal structure of mangrove patches. The study highlights the need for a hierarchical landscape approach when investigating animal-environment relations in marine landscapes.

No abstract