What was done
Recognizing that there is "debate about the relationship between temperature and widespread bleaching," and that "the exact physiological mechanism causing bleaching in corals is not known, nor is the manner in which temperature may affect this process," the authors examined a long (1966-1995) high-resolution sea surface temperature (SST) record from reef sites off southwestern Puerto Rico to determine its relationship to coral bleaching events.  They also evaluated nine temperature indices, derived from the SST data, for their ability to predict both moderate and severe bleaching events.

What was learned
Five of the temperature indices (maximum daily SST, and the number of days in which SSTs were above 29.5°C, 30°C, 30.5°C and 31°C) were able to correctly identify the four years in the record with severest bleaching.  However, the authors cautioned that these five indices "may be viewed best as probability based predictors rather than as causal conditions."

What it means
The authors suggest that their findings provide little support for the notion that the onset of bleaching is a function of cumulative heat stress, noting that "severe bleaching often occurred shortly after sharp temperature increases."  Rather, they believe that "prolonged heat stress may be an important precondition for bleaching to occur, with sharp temperature changes acting as an immediate trigger."