Johnston Atoll (JA) is an isolated reef in the north-central Pacific (16N,169W) and a military installation with a history of pollution. Environmental conditions have been strictly monitored since 1983; this was the first incidence of coral bleaching ever observed. It was also unusual because only corals within the lagoon were affected. No bleaching occurred at the reef edge.

CORAL BLEACHING ON JOHNSTON ATOLL, CENTRAL PACIFIC OCEAN
Anne L. Cohen, Phillip S. Lobel, and Gabrielle L. Tomasky

On 10 September 1996, extensive coral bleaching was noted on Johnston Atoll (JA), an isolated coral reef ecosystem in the central Pacific Ocean (16°N, 169°W). Between September 1996 and March 1997 we monitored the nature and extent of the bleaching, as well as the anomalous conditions of ocean temperature.

Coral "bleaching," or the loss of zooxanthellae and their photosynthetic pigments, is one of the first visible signs of thermal stress (1). The association between mass reef bleaching, and subsequent coral mortality, with elevated ocean temperatures is of concern in light of predicted global temperature increases over the next century (2). Mass bleaching is most often associated with anomalous ocean temperatures during the warmest month of the year. A temperature increase of 1°-2°C above the historical mean summer maximum is considered necessary to induce coral bleaching in tropical and subtropical environments (1).

The 1996 JA bleaching event did not occur in isolation but appears to have been part of a global-scale bleaching episode that began in the western Caribbean and Gulf of Mexico in the summer of 1995 and was observed at several sites in the western and central Pacific the following year (3) (Fig. 1a). Satellite images indicate a basin-wide sea surface temperature anomaly (SSTA) of between 0.5°C and 1.5°C in September 196 (4), coincident with reports of coral bleaching on the Hawaiian Island and on JA (Fig 1a).

On JA, we examined six lagoonal and reef-edge sites to depths of 5 m (Fig. 1b) during three field excursions: 2-4 October 21 October-5 November 1996, and 4-16 March 1997. Affected corals were tagged and photographed at one lagoonal site to monitor recovery rates. We distinguished between colonies that were fully bleached and those that showed partial bleaching, ie., bleaching confined to localized regions on a single colony. Our observations were as follows:

1. Bleaching was confined to corals in lagoonal sites (Fig. 1b). No bleaching occurred along the emergent reef with the exception of one bleached colony (Pocillopora meandrina) noted on the inside of the eastern reef edge.
2. Bleaching was species specific. All Montipora spp. and Pocillopora spp. were affected, although the degree of bleaching of individual colonies varied from completely unaffected to partially bleached to complete loss of skeletal pigment. Beaching was not observed in Acropora cytherea, the dominant coral species on JA.
3. Tissue loss from affected colonies did not occur during the first 3 weeks after the bleaching was noted, ie, 10 September-4 October 1996. Tissue loss was noted for several bleached Picillopora colonies in late October. On the contrary, bleached Monipora colonies maintained living polyps for the duration of the bleaching event.
4. We estimated the areal extent of the bleaching in lagoonal sites to be from 15% to 20% between 0 to 5 m depth. Bleached corals were observed to a depth of 5 m.
5. By March 1997, 50% of the affected, tagged colonies had made a full recovery and regained pigment. Recovery was unrelated to the degree of bleaching experienced by individual colonies but was to some degree species-specific in that Pocillopora colonies without tissue were eventually overgrown by algae.

Daily temperatures at two lagoonal sites and one reef-edge site were recorded using temperature loggers (Brancker Instruments) (Fig. 1b, c, d). The reef-edge site is a 6-m-deep channel (referred to as Munsens Gap; Fig. lb) in the emergent reef structure. We regard temperatures recorded at this site to be representative of open-ocean mixed layer temperatures. In-situ temperature records in combination with IGOSS NMC satellite- derived SSTs (4) (Fig. 1e) allow us to make the following observations:

1. Temperature loggers recorded a maximum summer lagoon temperature of 31.1°C on 25 August 1996, compared with a maximum summer temperature of 29.7°C in the previous year (14 September 1995) (Fig. 1c).
2. Average daily temperatures at the reef edge were 0.2°C lower than those recorded in the lagoon between 3 July and 21 October 1996. The maximum recorded SST at Munsens Gap was 29.8°C, and the maximum recorded difference between reef-edge and lagoonal sites was 0.4°C in late August 1996 (Fig. 1d).
3. Satellite-derived summer (JAS) SSTAs for a 1°x 1°C grid square centered on 16.5°N, 169.5°W indicate an anomaly of 0.6îC in 1996 (compared with the historical mean since 1982) (Fig. 1e).

In timing, nature, and extent-including the species affected, tissue loss, and rate of recovery-the bleaching episode on JA was very similar to the one that was observed on the Hawaiian Islands in the same year (3; Paul Jokiel, University of Hawaii, pers. comm.) and that was predicted on the basis of laboratory manipulations of temperature (5, 6). However, whereas temperatures between 28ƒC and 29ƒC are sufficient to induce bleaching of Montipora spp. and Pocillopora spp. on Hawaii (l), congenerics on JA appear tolerant of temperatures up to 29.8°C. Although the exact date when bleaching first occurred was not documented, the distribution of bleaching across the atoll, the timing of SST maxima in the lagoon, and the timing of the first observation of extensive bleaching lead us to conclude that the most sensitive coral species on JA have an upper thermal limit of about 30ƒC. This apparent difference in coral thermal tolerances between Hawaii and JA corresponds to differences in the maximum summer temperatures between these two sites.

Although in situ temperature data enable us to estimate the upper thermal limit of the affected JA coral species, the time series are too short to determine how high this limit is above normal summer SST maxima. The longer satellite-derived temperature record shows an anomaly of 0.6°C during the summer of 1996; according to field and laboratory observations, this anomaly is not high enough to induce coral bleaching by temperature alone (1,5,6). However, in assessing whether temperature was the sole cause of the JA coral bleaching event or whether other factors were involved, it is important to recognize the spatial resolution over which satellite-derived SSTs are aver- aged (1° X 1°). Satellite temperatures are therefore representative of relatively large-scale open-ocean conditions and do not distinguish fine-scale temperature variability across the atoll. Our logger data indicate that small but important differences in SST occurred between sites on JA. SSTs in the lagoon, where bleaching occurred, were up to 0.4æC higher than those at the reef edge, where bleaching did not occur. Thus, we deduce that lagoonal temperatures were at least 1æC higher than the long- term ambient summer SST, which is in good agreement with that predicted to induce coral bleaching (1, 5, 6).

This study was supported by the Office of Naval Research grant #'s: NOOO-14-91-J 1591, N00014-92-J 1969 and N00014- 95 I- 1324, with additional support from the US Army Chemical Demilitarization Program, the U.S.A.F., the U.S.C.G., and the U.S. Army Legacy Program DAMD 17-93-53052. We thank David Shogrun and Gary McCloskey for field assistance and two anonymous reviewers for helpful comments on the original manuscript.

Literature Cited

1. Jokiel, P. L., and S. L. Coles. 1990. Coral Reefs 8: 155- 162.
   
2. Williams, E. H., and L. Bunckley-Williams. 1988. Workshop Report, The National Undersea Research Program, NOAA/OAR, Silver Spring, Maryland.
3. NOAA Coral Health and Monitoring Program website address (aoml.noaa.gov/). National Oceanic and Atmospheric Administration.
4. LDEO Climate Data Catalog website address (http://ingrid.ldgo.columbia.edu/). Lamont-Doherty Geological Observatory, Columbia University.
5. Coles, S. L., P. L. Jokiel, and C. R. Lewis. 1976. Pac. Sci. 30: 159-166.
6. Coles, S. L., and P. L. Jokiel. 1978. Mar. Biol. 49: l87- 195.