accumulations of cyanobacteria could cause local increases in seawater temperature of up to 1.5C. Similarly, Lewis et al. (1983) reported that the deep chlorophyll maximum in the eastern North Atlantic caused the development of a warm layer of water to develop below a colder layer. These studies demonstrate that attenuation of light in the water column by phototropic organisms, if they are present in sufficient numbers, can cause differential heating of the water mass, whereas in their absence this does not occur.
Blue Holes are cave systems that have developed within the Bahamian carbonate platform (Sealey 1994; Palmer 1985). The size of these cave systems can be can be extensive. Laterally cave passages can extend several kilometers and vertically blue holes may range in depth from 10 to several hundred m. These cave systems connect with the sea via the lateral passages. A more intriguing series of cave systems also found in the Bahamas are Black Holes, the most spectacular of which is the South Andros Black Hole which has an almost circular entrance 300 ± 15 m in diameter and is *47 m deep (Palmer 1985; Schwabe and Herbert 2004). Unlike Blue Holes, the Black Holes have no connection with the sea except via seepage and rock fractures, and fresh water sits on a layer of saline water with little or no mixing. From the air the water of the South Andros Black Hole appears dark blue/black (Fig. 1). In this article we show that the color of the water is due to a 1-m dense layer of purple sulfur purple bacteria located at depth of 17.8 m, which effectively absorbs all the incom- ing photons. At this depth the ambient water temperature increases sharply. Below the bacterial layer the water temperature cools rapidly. In this study, we describe the processes, which account for the observed increase in water temperature in the bacterial layer.
Fig. 1 Aerial photograph of the South Andros Black Hole cave system, the Bahamas. Reprinted from Quaternary International, 121, Stephanie Schwabe and Rodney A. Herbert, Black Holes of the Bahamas: What are they and why they are black, 3–11, Copyright (2004), with permission from Elsevier
Cell culture and harvesting
Thiocapsa BH-1 and Allochromatium BH-2 isolated from the South Andros Black Hole (Herbert et al. 2005) were grown photosynthetically in the mineral salts medium of Pfennig and Tru¨per (1992) supplemented with 3% w/v NaCl. Cells were harvested and membranes prepared according to the method described by Evans et al. (1990).
Determination of growth temperature profiles
Growth temperature profiles for Thiocapsa strain BH-1 and Allochromatium BH-2 were determined by incubating the cultures in the mineral salts growth medium of Pfennig and Tru¨per (1992) supplemented with 3% w/v NaCl. The mineral salts medium was dispensed in 150 · 15 mm diameter screw-top tubes (Corning Life Sciences) which were pre- equilibrated at the designated incubation temperature in thermostatted glass aquarium tanks set at the following temperatures: 5, 10, 15, 20, 25, 30, 35, and 40C ± 2C. After equilibration for 1 h each series of tubes was inocu- lated, respectively with 1 ml volumes of exponentially growing cultures of Thiocapsa BH-1 and Allochromatium BH-2. The inoculated tubes were replaced in the thermo- statted aquaria and continuously illuminated using tungsten bulbs to give a light intensity at the tube surface of 5,000 lux. At 24 h intervals the tubes were removed and vortexed to ensure the cells remained suspended and when appropriate the cultures were fed with neutralized sulfide. The optical density of the elemental sulfur depleted cultures was mea- sured at 650 nm after 7 days incubation, using a Bausch and Lomb Spectronic spectrophotometer.
Absorption spectra were collected at 77 K in an SMC-TBT flow cryostat (Air Liquide, Sassenage, Fr.) cooled with liquid nitrogen, or at room temperature, using a Varian Cary E5 Double-beam scanning spectrophotometer. Sam- ples for low temperature absorption measurements contained 60% (v/v) glycerol.
Room-temperature fluorescence excitation spectra were obtained with a SPEX Fluoromax spectrofluorimeter (ISA, Longjumeau, France) equipped with a red sensitive pho- tomultiplier R406 (Hamamatsu Hamamatsu Electronics, Japan). The efficiency of carotenoid (Car) to bacterio-
photosynthetic membranes was calculated not from the absorption spectrum, but from the fractional absorption