Technical Comments

Response to 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.1128502


Coral reefs are exceptional environments where changes in calcification, photosynthesis, and respiration induce large temporal variations of pH. We argue that boron isotopic variations in corals provide a robust proxy for paleo-pH which, together with the likely concomitant changes in the reconstructed partial pressure of CO2 (PCO2) calculated by Matear and McNeil, fall within ranges that are typical of modern coral reef ecosystems.

Matear and McNeil (1) recognize the usefulness of boron isotopes in corals to record changes in seawater pH but question our pH reconstruction (2) for three reasons. First, they suggest that the reconstructed pH variability (7.9 to 8.2 units) is unreasonably large, because it would imply variations in PCO2 and alkalinity that are unrealistic. Second, they question whether such large changes in PCO2 could have been maintained for decades. Third, they argue that had these PCO2 changes persisted, that large variability in the air-sea flux of CO2 would have altered the d13C-DIC (dissolved inorganic carbon) values and that this should be reflected in the coral skeleton d13C.

As stated in (2), the pH variability of our reconstruction represents Flinders Reef seawater, not the open ocean. Local pH values can change considerably, especially within coral reefs where changes in calcification, photosynthesis, and respiration have been shown to induce large variations of pH over diurnal and seasonal time scales. Yates and Halley (3) reported diurnal pH changes from 7.82 to 8.42 units in a Molokai reef flat, with associated ambient seawater PCO2 values of 170 to 935 matm. Ohde and van Woesik (4) found diurnal changes of up to 0.7 pH units in a coral reef atoll close to Okinawa, with concomitant changes in PCO2 of 100 to 900 matm. Suzuki et al. (5) reported diurnal variations of up to 1 pH for a stagnant coastal reef, and Schmalz and Swanson (6) reported diurnal changes of about 0.15 pH units for the Enewetak atoll, an open-ocean reef similar to Flinders Reef. This latter variation is similar to the seasonally resolved change observed in Flinders Reef, based on our high-resolution coral d11B data [figure 2C in (2)].

Matear and McNeil (1) estimated associated PCO2 changes in Flinders Reef water using the assumption of constant alkalinity and obtained variations between 250 and 590 parts per million, which they considered unrealistic. However, if the buildup of reef-water CO2 due to calcification is a main mechanism for pH variation, alkalinity will decrease to some extent during the low reef-water pH periods, rendering the assumption of constant alkalinity invalid. Nevertheless, even this wide range of PCO2 values falls within the range commonly encountered in reef systems (37). In addition to calcification, the buildup of CO2 from respiration will play a key role in decreasing pH during periods of poor flushing of reef waters. These factors, combined with possible oceanographic changes in the open-ocean water masses bathing the reef, hinder the calculation of the whole set of seawater carbonate system parameters based on boron isotope–inferred pH data alone.

Matear and McNeil also claim that the high levels of PCO2 in Flinders Reef could not be maintained for decades assuming an air-sea equilibration rate of less than 1 year for CO2. We agree that reef-water pH at Flinders is likely to vary on shorter time scales, but the 5-year resolution of the coral paleo-pH record averages any diurnal to seasonal variability. Even during high-PCO2 (low pH) multidecadal periods, PCO2 in Flinders reef water may have periodically reached lower values closer to global atmospheric levels. On the other hand, coral reef systems that are not dominated by algae, like Flinders Reef, are known to be sources of atmospheric CO2 because the release of CO2 from coral calcification is greater than that fixed by photosynthesis (8, 9). Therefore, we maintain that the effects of progressive acidification of the oceans are likely to differ between coral reefs because reef-water PCO2 and consequent changes in seawater pH will rarely be in equilibrium with the atmosphere.

Finally, Matear and McNeil note that large changes in the air-sea flux of CO2 would alter the d13C-DIC because the air-sea flux process fractionates 12CO2 preferentially relative to 13CO2. However, the air-sea equilibration time for carbon isotopes is about 10 years (10), which is 10 times as long as for CO2 itself. Furthermore, coral skeletal d13C has proven to be a complex nonunique environmental tracer because of complicated interactions that involve strong isotopic fractionation including photosynthesis, respiration, skeletal calcification, and the changing balance between heterotrophy and autotrophy (e.g., 1115). The sum of these influences on coral d13C could easily mask the effect of any change in d13C of DIC brought about by air-sea fluxes of CO2.

In summary, the arguments proposed by Matear and McNeil are based exclusively on open-seawater behavior of carbon system parameters. By contrast, our reconstruction focuses on seawater properties of a coral reef, where local processes can induce large variations in pH. Our study (2) clearly demonstrated the potential of boron isotopes in massive coral skeletons to contribute to our understanding of how the world's coral reefs will respond to future ocean acidification. Given that instrumental records of seawater pH exceeding a single decade are not yet available, the boron isotopic composition of long-lived corals currently offers the only practical means to determine such changes back through time.

References and Notes

Stay Connected to Science


Navigate This Article