MEPS 247:183-195 (2003)  -  doi:10.3354/meps247183

ABSTRACT: On naturally occurring seagrass patches, 10 plots of 100 m² each were permanently marked. Seagrasses, macroalgal components and fish assemblages were monitored in February, April and June 1996. Maps of the patch configurations were produced in January 1996 by photographing components and were described in terms of the area of seagrass in a plot (cover), the average size of each seagrass patch, the total number of seagrass patches in a plot, number of halos, shape of a patch (perimeter:area ratio), and complexity of the edges (fractal dimension). Extraction of principal components axes (PCA) on variables describing patch and macrophyte characteristics was employed. Fish numbers from the patchy plots standardized against the fish numbers caught on the corresponding nearby continuous bed showed a positive parabolic relationship against the PCA interpreted as Œcontinuity of vegetation¹ for the months of February and June; where curvilinear regressions gave values of r = 0.85 and 0.94, respectively. In April, a negative parabola was exhibited against the PCA interpreted as Œmacrophyte stands within a patch¹ (r = 0.73). This trend was basically due to the response of the fishes Siganus fuscescens and Abudefduf sexfasciatus dominating the catch and their preference to intermediate densities of seagrass as the optimal refuge. On the other hand, positive parabolic trends exhibited in February and June imply that fishes tended to prefer plots, which were either extremely patchy or highly continuous, while plots with spatial configurations of average nature were avoided. The differences in fish assemblages and/or seagrass densities within the vegetation for the sampling months explain the different responses. Management implications are discussed.

KEY WORDS: Seagrass · Patchiness · Fish · Landscape ecology