162 



MARSH 



average rate of 4000 g CaCOa m^^ yr^'. There were no 

 apparent differences between day and nigfit in these 

 studies. Smith further calculated that, although there was 

 little particulate CaC03 removal from the reef flat over the 

 duration of his studies, there has been virtually no net 

 CaCOs accumulation on the windward reef flats of 

 Enewetak over the last several thousand years. He thought 

 that lagoonward accumulation is the probable sink for cal- 

 careous material produced on the reef flat but that sedi- 

 ment transport occurs almost exclusively during periods of 

 intense wave action. 



The observation that daytime and nighttime calcifica- 

 tion rates for the whole community are similar ran counter 

 to much previous thinking, which was strongly influenced 

 by the measurements of individual organisms enclosed in 

 small containers. For instance, Goreau (1961) reported 

 that calcification rates in individual corals, as measured by 

 uptake of ''^Ca, were strongly light-dep>endent. However, 

 Smith pointed out, as had others before him (e.g., Chave 

 et al., 1972), that there are large uncertainties inherent in 

 using the standing crop and turnover of individual organ- 

 isms to estimate CaCOs production of the community as a 

 whole. Furthermore, it is likely that corals, which are most 

 often used in the individual-organism approach, account for 

 a minor component of total calcium carbonate production 

 on a reef. Smith also pointed out that the technique of 

 measuring alkalinity depletion as a way to estimate CaCOa 

 deposition could be applied in incubation chambers with 

 individual organisms. Smith (1974) stated in a later paper 

 that community precipitation of CaCOs, ranging from 

 — 0.02 to 0.2 moles CaC03 m~^ d"\ is an order of 

 magitude lower than the calculated CO2 flux resulting from 

 organic carbon metabolism (±0.2 to 6 moles d^'), 

 although the former process is not reversible on a 

 day-night cycle and the latter process is. Flux due to diffu- 

 sion across the air-sea interface or to mixing of water 

 masses is likewise an order of magnitude lower than that 

 resulting from organic carbon metabolism. 



Smith and Kinsey (1976) combined the data and ideas 

 generated by the Enewetak research with those from reef 

 research elsewhere to make some generalizations about 

 calcium carbonate production and sea level change. They 

 suggested that shallow, seaward portions of most modern 

 reefs produce approximately 4 kg CaCOs m~^ yr~\ and 

 that "protected" areas produce about 0.8 kg. They argued 

 that the difference in these rates is probably due largely to 

 differences in water motion. The more rapid rate is 

 equivalent to a maximum vertical accretion of 3 to 5 mm 

 per year and places an upper limit on the potential of 

 modern coral reef communities to create a significant verti- 

 cal structure during a rising sea level. They suggested that 

 the major taxa accounting for most CaC03 deposition 

 rates are corals, coralline red algae, and calcareous green 

 algae; the potential reef accretion, however, does not 

 appear to be affected by coral versus algal dominance. 

 They found little evidence for latitudinal gradients. 



Smith and Harrison (1977), using dome enclosures 

 placed over the benthic community and following pH and 



alkalinity changes in the enclosed water mass, assessed cal- 

 cium carbonate production of the mare incognitum on the 

 upper seaward reef slope, a habitat barely considered in 

 any previous study of any type. They made measurements 

 on "vasiform" Acropora heads and on algal pavement but 

 not on sand and rubble substrata. Smith and Harrison 

 reported that calcium carbonate production by the corals is 

 substantially lower on the seaward slop>c than at the con- 

 trol site (a subtidal quarry on the reef flat of Enewetak 

 Island) and that it may decrease with depth. Production by 

 algal pavement was also reported to be lower on the slope 

 than at the control site but showed no apparent reduction 

 with depth. However, calcium carbonate production by 

 algal pavement in the quarry was dramatically slower than 

 that by algal pavement on the reef crest. The coral calcifi- 

 cation rate (on a square-meter basis) was always greater 

 than that for algal pavement by a factor of 3 to 9. As in 

 other reef habitats, it was recognized that topographic in- 

 equalities of the mare incognitum increase the effective sur- 

 face area by up to 50%. Smith and Harrison concluded 

 that the most actively calcifying portion of an atoll is near 

 sea level, even though standing crops of calcifying organ- 

 isms on the reef flat may be lower than on the reef slope. 

 This is consistent with assumptions inherent in earlier 

 studies that the major metabolic activity is on the reef flat 

 rather than in other subsystems of the atoll. Their 

 presumed optimum environment for reef development is a 

 broad shoal area only a few meters deep with exposure to 

 oceanic swells. 



Nitrogen Flux 



It was Odum and Odum (1955) again who first con- 

 sidered nitrogen flux on reefs and made a few measure- 

 ments of changes in nitrate and ammonium as waters 

 flowed across the windward reef flat. Gilmartin (1960) like- 

 wise made a few measurements in lagoon waters by "stan- 

 dard oceanographic techniques," but the first extensive 

 measurements of any nitrogen compounds were made dur- 

 ing Project SYMBIOS (Johannes et al., 1972; Webb et al., 

 1975). The researchers observed changes in various nitro- 

 gen species as water flowed across the usual coralgal or 

 exclusively algal transect and found that both transects 

 showed a significant net export of combined nitrogen, 

 implying a large input of nitrogen into the system from a 

 source other than combined nitrogen in incoming waters. 

 Following up on this observation, they found that there 

 were high rates of gaseous nitrogen fixation in the reef 

 ecosystem, the first time that this process had been 

 reported for any such system. The transect dominated by 

 algae showed a net uptake of nitrate-nitrogen, and there 

 was a net export of that species from the coralgal transect. 

 Other nitrogen species (ammonium, dissolved organic nitro- 

 gen [DON], and particulate organic nitrogen [PON]) like- 

 wise increased significantly in waters flowing across this 

 transect. The DON concentrations (2300 to 3000 nmol) 

 were about an order of magnitude higher than the PON 

 values. On this transect, there was a net removal of N03~ 



