Ecological patterns and processes are often scale-dependent. Understanding organisms' perceptions of and responses to landscape heterogeneity and connectivity is essential for effective conservation and management. We used multiple regression models with backward elimination to test relationships between juvenile fish density and diversity, quantified visually by snorkelers, and site-(100s m) and landscape-scale (1 km) habitat characteristics of mangroves in the Great White Heron National

more »

... fe Refuge (GWHNWR) in the lower Florida Keys, USA. We compared site and landscape model performance using Akaike's information criterion (AIC) and adjusted R 2 values. Our results demonstrate the following: (1) juvenile fish density in mangrove backreef habitats correlated with landscape characteristics; (2) relationships between fish density and site or landscape characteristics differed greatly among species; and (3) juvenile fish diversity was not strongly correlated to either site or landscape characteristics. We therefore advise caution when (1) selecting fish species and sites for experiments to test nursery habitat hypotheses given the scale-and species-specific relationships between fish and mangrove habitats observed in the present study, and (2) implementing conservation strategies based upon habitat surrogates or quotas (e.g. protect 20% of mangrove habitat to conserve fish species) because landscape context might strongly, and uniquely, influence individual fish species' juvenile density. In light of our results, we argue in support of landscape analyses and individual-based modeling as useful tools to prioritize conservation of backreef nursery habitats. KEY WORDS: Backreef habitats · Nursery role · Habitat quality · Spatial scale · Conservation · Coral reef fish · Landscape ecology Resale or republication not permitted without written consent of the publisher Editorial responsibility: Kenneth Heck,