Abstract

Accurately assessing reef fish community assemblages is an important feat for ecosystem health, fisheries regulations, and invasive species monitoring. However, several underwater surveying techniques exist to do this, each with benefits and drawbacks. Reef visual censuses (RVCs), or Standard visual censuses (SVCs), are the most efficient and commonly used but are thought to underestimate species densities compared to belt transect and species-specific roving surveys. We compared density estimates across reef fish species for all three survey types using data from the Florida Keys region to determine SVC performance and looked at these differences in relation to predictors such as habitat traits, survey traits, and species traits using linear mixed effects models. We additionally compared density and frequency of occurrence estimations across survey methods for all recorded species and SVC focal species and invasive lionfish individually. Variation was found in SVC performance compared to transect and roving surveys, with transect surveys recording higher densities across most species and roving surveys recording lower average densities and higher frequencies of occurrence for most, including two SVC focal species (red and black grouper) and lionfish. Several traits were significant predictors of density differences, including commonalities such as poorer SVC performance for cryptic species. Our results suggest variable SVC performance depending on the species and traits, and we caution managers to take species and habitat traits into account when designing accurate survey techniques

PDF Preview