Technical Comments

Comment on "Preindustrial to Modern Interdecadal Variability in Coral Reef pH"

See allHide authors and affiliations

Science  27 Oct 2006:
Vol. 314, Issue 5799, pp. 595
DOI: 10.1126/science.1128198


Based on the boron isotopic composition of coral from the southwestern Pacific, Pelejero et al. (Reports, 30 September 2005, p. 2204) suggested that natural variations in pH can modulate the impact of ocean acidification on coral reef ecosystems. We show that this claim cannot be reconciled with other marine carbon chemistry constraints and highlight problems with the authors' interpretation of the paleontologic data.

Pelejero et al. (1) analyzed the boron isotope composition of Porites coral from the southwestern Pacific and reported large interdecadal variations in pH (between 7.9 and 8.2) with a strong periodicity of about 50 years. This implies that natural pH variability in the ocean is greater than pH variation attributed to anthropogenic factors (a decrease of 0.1 pH units since 1880). (2). Pelejero et al. argue that the implied natural pH variability cannot be associated with variations in either sea surface temperature (SST) or sea surface salinity (SSS). Instead, they attribute the pH variation to “coral reef calcification combined with limited flushing of reef water exerting an important local control over the buildup of partial pressure of CO2 (PCO2) within the reef.” However, they do not calculate the implied PCO2 magnitude associated with such a large pH change. If we use the observed water properties (SST = 25°C, SSS = 35.3) and the authors' assumption of constant alkalinity (2300 μmol/kg), the range in pH implied from their boron isotope measurements must be associated with interdecadal PCO2 variations between 250 and 600 parts per million (ppm) (Fig. 1, points A and B). Because preindustrial atmospheric PCO2 values were about 280 ppm, the inferred estimate of pH range implies that PCO2 within the seawater of the coral reef was maintained at levels up to double that of the atmosphere for a decade. However, this interpretation is unrealistic for two reasons. First, the equilibration rate of CO2 for the ocean is less than 1 year, and even shorter for very shallow reef environments where the observations in (1) were collected. The implication that PCO2 can build up over 300 ppm beyond the atmospheric concentration on a decadal time scale is not possible because exchanges of CO2 with the atmosphere would bring the PCO2 of the water back to a value that is in equilibrium with the atmosphere. Second, if PCO2 were to experience such large variations as suggested by the pH changes, this would drive large changes in air-sea flux of CO2. Such changes would in turn alter the isotopic composition of dissolved inorganic carbon (δ13C-DIC) because the air-sea flux process fractionates 12CO2 preferentially relative to 13CO2 (3). Therefore, we would expect to find clear correspondence between the interdecadal variability in the calculated pH and the observed δ13C-DIC, which is not the case.

Fig. 1.

Contour plot of alkalinity (μmol/kg) as a function of pH and PCO2. Points A to B denote the trajectory for reproducing the observed pH change but maintaining a constant alkalinity of 2300 μmol/kg as assumed in (1). Points A to C denote the trajectory for reproducing the observed pH change but maintaining constant PCO2 as required to satisfy the physical constraint that the seawater would equilibrate with the atmosphere in less than 1 year. The figure demonstrates that an alkalinity change from 2300 μmol/kg to 1000 μmol/kg is required to reproduce the observed pH change but maintain a PCO2 value close to the initial value at point A.

Alkalinity changes could potentially account for the large pH variations reported by Pelejero et al. However, to satisfy the implied pH variations would require an interdecadal change in alkalinity from 2300 μmol/kg to 1000 μmol/kg (Fig. 1, points A and C). These alkalinity changes are unreasonably large given that the typical variations within the studied section of the Great Barrier Reef are only about 40 μmol/kg (4, 5). Other studies on coral reefs in the Indo-Pacific show maximum alkalinity variability of 100 μ mol/kg (6, 7), far less than the 1300 μ mol/kg that is required over decadal periods in the Pelejero et al. study. Calcification is the main process that reduces alkalinity of seawater around the reef, but the calcification rates of the corals used to estimate the pH in (1) do not show any correlation with the pH variations. Further, empirical studies of coral reef calcification show declines in calcification as the pH declines (8), which is counter to the trend required to produce the inferred pH variability. Therefore, explaining the large pH variability in terms of changes in alkalinity also appears unrealistic.

Ocean acidification is clearly a threat to coral reef ecosystems, and the implication of the Pelejero et al. study is that these ecosystems are subject to large natural variations in pH. This interpretation has important implications for assessing the response of coral reefs to future ocean acidification. Although boron isotopes have been shown to be a useful pH proxy, our carbon chemistry calculations show that the Pelejero et al. interpretation of the paleontologic data suggesting large decadal changes in pH is not consistent with acceptable changes in either PCO2 or alkalinity. This ultimately casts doubt on the notion that corals have already undergone such large natural variations in pH and are capable of adapting to the pH changes projected to occur over this century.

References and Notes

View Abstract

Navigate This Article