Methods for monitoring the status of marine communities are increasingly adopting the use of images captured in the field. However, it is not always clear how data collected from photographic images relate to historic data collected using traditional underwater visual census methods. Here, we compare coral health and disease data collected in situ by scuba divers with photographic images collected simultaneously at 12 coral reef sites. Five globally relevant coral diseases were detected on 194 colonies from in situ surveys and 79 colonies from photos, whilst 698 colonies from in situ surveys and 535 colonies from photos exhibited signs of compromised health other than disease. Comparisons of in situ surveys with photographic analyses indicated that the number of disease cases occurring in the examined coral populations (prevalence) was six times higher (4.5 vs. 0.8% of colonies), whilst compromised health was three times higher (14 vs. 4% of colonies) from in situ surveys. Skeletal eroding band disease, sponge overgrowth and presence of Waminoa flatworms were not detected in photographs, though they were identified in situ. Estimates of black band disease and abnormally pigmented coral tissues were similar between the two methods. Estimates of the bleached and healthy colonies were also similar between methods and photographic analyses were a strong predictor of bleached (r 2 = 0.8) and healthy (r 2 = 0.5) colony prevalence from in situ surveys. Moreover, when data on disease and compromised health states resulting in white or pale coral colony appearance were pooled, the prevalence of 'white' colonies from in situ (14%) and photographic analyses (11%) were statistically similar. Our results indicate that information on coral disease and health collected by in situ surveys and photographic analyses are not directly comparable, with in situ surveys generally providing higher estimates of prevalence and greater ability to identify some diseases and compromised states. Careful sampling of photographs can however identify signs of coral stress, including some coral diseases, which may be used to trigger early-warning management interventions.
Keywords: Bleaching; Coral disease; Coral reef monitoring methods; Methods comparison; Underwater visual census.