Photographic versus visual census techniques for the quantification of juvenile corals

The abundance of juvenile corals was quantified by visual and photographic methods to determine whether photography could be used as a rapid census technique in reef monitoring projects. Between 9 and 17 0.25-m 2 quadrats were positioned randomly along replicate 25-m transects on reefs in St. John (U.S. Virgin Islands), the Florida Keys, and Belize. In the visual method, all juvenile corals (0.4 cm ≤diameter ≤5.0cm) in both open and cryptic locations within the quadrats were identified to genus and counted. A close-up photograph of a 0.039-m 2 planar area was then taken in the center of each quadrat for subsequent analysis. Plots of the running mean and standard error of juvenile density against the number of transects produced asymptotes after approximately three transects at each site, suggesting that sample sizes were sufficient for both methods. However, juvenile densities estimated by the visual method were not correlated with values obtained by the photographic method at any site. These discrepancies cannot be attributed to dissimilar sampling areas, because analysis of variance demonstrated that the difference between methods varied among sites, and because similar variation was apparent when visual and photographic surveys were compared for similar sample areas. Rather, the discrepancies are a consequence of juvenile corals growing in microhabitats where they cannot be quantified in planar photographs. We recommend the visual technique to quantify juvenile corals; at four sites we found that adequate estimates of juvenile density can be obtained from surveys of 0.5 x 0.5 m quadrats, randomly positioned at 17 locations along each of four 25-m transect lines.