One of the main issues regarding artificial reef fish evaluation is the performance of an efficient visual census sampling. An experimental artificial reef made of four different types of concrete modules was implanted on the north coast of Rio de Janeiro (21º29'S, 41º00W). Two interval-sampling periods using a visual census method were performed in the summer of 2003 (N = 6) and 2004 (N = 6), weekly and daily, respectively. The longer interval-sampling period showed twice of species with eight exclusive ones. Total fish abundance was higher during the shorter interval-sampling period, due to the presence of large shoals of Chloroscombrus chrysurus (Linnaeus, 1766) and Haemulon aurolineatum (Cuvier, 1829) especially at the complex modules. Fish abundance according to vertical position and feeding habits in the two interval sampling-periods showed a clear association to the structural complex modules, reinforcing the shelter influence on local fish assemblage structure and composition. The second survey showed quite different diversity and taxonomic results and highlighted the impossibility of obtaining several independent samples in a short period.of time.
Experimental reef complex; Rio de Janeiro; sampling period; underwater visual census
