Coefficient of variation of sea surface temperature (SST) as an indicator of coral bleaching
Sammarco, P.W.; Winter, A.; Stewart, J.C. (2006). Coefficient of variation of sea surface temperature (SST) as an indicator of coral bleaching. Mar. Biol. (Berl.) 149(6): 1337-1344. http://dx.doi.org/10.1007/s00227-006-0318-0
In: Marine Biology: International Journal on Life in Oceans and Coastal Waters. Springer: Heidelberg; Berlin. ISSN 0025-3162; e-ISSN 1432-1793, more
|  |
| Authors | | Top |
- Sammarco, P.W.
- Winter, A.
- Stewart, J.C.
|
|
|
| Abstract |
Coral bleaching has become a major problem on reefs around the world in recent decades. It is believed that mean temperature alone is the primary force driving this ecological phenomenon. We propose that variance in temperature in the short term is just as important as the mean. Thirty years of daily sea surface temperature (SST) data have been collected by the University of Puerto Rico at Mayaguez Marine Laboratory in La Parguera, PR. These data were collated and analyzed initially (by Amos Winter) for their relationship to coral bleaching in this area. We found that the data fell into three categories: high mean temperatures associated with severe bleaching, cooler mean temperatures associated with no bleaching, and years of high SSTs but with no coral bleaching. Here, we examined the relationship between mean temperature during those months in which bleaching occurred, temperature variance (as measured by standard deviation), and coefficient of variation (CV; i.e., SD standardized by the mean). We also derived a critical threshold temperature and level of resolution in time for calculating these statistics to clearly describe the circumstances of bleaching versus non-bleaching events, particularly at marginal bleaching temperatures. These characteristics were compared for the four warmest months of the year (July–October) for four warm bleaching years (1969, 1987, 1990, and 1995), four cool non-bleaching years (1984, 1985, 1986, and 1988), and two warm non-bleaching years (1994 and 2000). No relationship was found between the mean SST and SD in terms of predicting bleaching. The two primary statistics which, in concert, did indicate bleaching, however, were the short-term, biweekly mean temperature and its the associated CV. Bleaching occurs in association with both high temperatures and a high CV. The CV becomes a critical determinant of bleaching only when temperatures are ~29.1–29.8°C. The warm, non-bleaching years were generally characterized by a CV of < 1.9 and a temperature range between 28.5 and 29.9°C. We conclude that increased mean SSTs alone are not sufficient to induce coral bleaching; a high variance in SST at marginal, lower bleaching temperatures can induce bleaching, and likewise, a low variance of such will not induce bleaching. This variance is most clearly described by the CV. |
|
