A Comparison of the Diagnostic Accuracy of in-situ and Digital Image-Based Assessments of Coral Health and Disease

Abstract

The prevalence of coral disease is steadily increasing throughout the global ocean, and there is a growing need for efficient methods for detecting and monitoring coral health. At present, coral health assessments are primarily conducted using in-situ surveys, which record visual observations of disease in the field. Recent technological advancements allow researchers to instead collect high-resolution imagery of benthic habitats, and these images can be used in conjunction with digital tools to assess the health of coral colonies at a later time. However, little is known about the relative efficacy or diagnostic accuracy of these two approaches. This study contrasts the diagnostic accuracy of in-situ and digital methodologies for detecting diseases and adverse health conditions affecting corals. Multiple 1 m2 plots are surveyed on coral reefs located on both the windward and leeward side of Hawaii Island. For each plot, an in-situ visual analysis of coral health is conducted by a diver and images are collected and rendered into a high-resolution orthomosaic for subsequent digital analysis. Both methods assess the same coral colonies, resulting in paired health diagnoses for multiple health conditions. Lacking a gold-standard diagnosis of health conditions, a latent class model is used to estimate the sensitivity (true positive rate) and specificity (true negative rate) of both methods. We find that in-situ assessments of coral health have a higher sensitivity and lower specificity in detecting health conditions when compared to digital analyses based on orthomosaics. However, the effect size is relatively modest, indicating that while the in-situ method provides a more sensitive diagnostic approach, the techniques are of comparable accuracy, and should both be considered viable methods of characterizing and monitoring coral health.