BULLETIN OF MARINE SCIENCE, 69(2): 487–496, 2001 487 A PROPOSED METHOD FOR CORAL COVER ASSESSMENT: A CASE STUDY IN ABROLHOS, BRAZIL Bárbara Segal and Clovis B. Castro ABSTRACT One of the methods most often used to evaluate reef community coverage are line intercept transects (LIT). However, collection of cover data using LIT has drawbacks. The optimization of data collecting permits an increase in number of replicates, allowing a more adequate quantification of the community and comparison among different sites/ times. It was assumed that a LIT is composed of several adjacent data points and the number of data points necessary to ‘describe’ a whole line was investigated. Several LIT were deployed in the Abrolhos Archipelago, Brazil. The variation of the stand error for percent cover in relation to a varying number of points drawn from each LIT was evalu- ated. A relatively stable condition was reached with 500 (out of 2000) points. This proce- dure (point intercept transects - PIT) needed much less effort for two reasons: (1) time per transect (up to 62.5% reduction); and (2) for each point the diver can just mark spe- cies/categories in a ticklist, so less underwater skills are needed. We also demonstrate and discuss the necessity for line replicates. Species with less than 2% cover were detected in LIT and PIT. The Abrolhos reefs, situated off the eastern coast of Brazil, are the most developed reef assemblages of the South Atlantic Ocean (Leão and Ginsburg, 1997). These reefs present some features, such as geomorphology, low coral diversity, high degree of coral ende- mism, and depositional setting (Laborel, 1970; Leão, 1996). The reefs are composed of isolated or fused reef columns, named chapeirões (Leão et al., 1997). In spite of their ecological importance, few quantitative benthic surveys have been conducted in the reefs of the area (but see Pitombo et al., 1988; Coutinho et al., 1993; Villaça and Pitombo, 1997; Figueiredo, 1997). There are several constraints to sampling benthic reef communities. These include ob- server effort and skills needed to collect and record data underwater, distance from main- land, and problems related to weather conditions (boating and diving). Also, travel, boat- ing, and equipment costs make it difficult to carry out large-scale sampling. This is com- pounded by logistic constraints or in areas of difficult access, as is the case for many Brazilian reefs. Because funding is limited, and cost is related to time and effort in the field, sampling replication is often restricted by these features. This condition may lead to samples that are not representative of the area under study. In Brazilian reefs the need for adequate sampling methodology is greater still because of the low coral coverage and species diversity, the small size of most coral species, and high water turbidity. This ex- cludes the use of some of the methods developed in other locations such as video and photographic transects. These features can also demand a larger number of replicates to adequately describe an area, as low coverage or small species are often misrepresented with a small number of replicates. Sampling design, however, must balance costs, effort, and adequacy. An adequate sampling design includes an appropriate choice of sample size (including replication). If sample size is too small, the power of statistical tests may be insufficient to test hypotheses involving small differences between means (Bros and