Comment on “empirical Partition Coefficients for Sr and Ca in Marine Barite: Implications for Reconstructing Seawater Sr and Ca Concentrations” by Kristen B. Averyt and Adina Paytan

Dayong Wang,Xiancai Lu,Shijin Xu

DOI: https://doi.org/10.1029/2006gc001456

2007-01-01

Abstract:[1] Averyt and Paytan [2003] analyzed marine barite sampled from the modern sediments of the southern Ocean and the Pacific Ocean. On the basis of the average value of Sr/Ba ratios of all these barite samples ([Sr/Ba]barite), they deduced an empirical partition coefficient for Sr in marine barite (DSr) to reconstruct the Sr concentration of the ancient oceans. Following the methodology and the formula described by Averyt and Paytan [2003], we can conclude that the availability of this DSr depends on the prerequisites as follows: (1) The Sr concentration of seawater ([Sr]sw) is practically homogenous in world oceans (with the average value of 87.4 μM (±0.56%)) [de Villiers, 1999; Averyt and Paytan, 2003], and [Ba]sw is the identical saturation concentration of Ba in seawater at the approximate depth of barite formation (i.e., [Ba]ss, 80 nM [Monnin et al., 1999; Averyt and Paytan, 2003]); (2) Sr incorporates into marine barite in the proportion to the ratio of [Sr]sw divided by [Ba]ss, and this proportion should be almost identical for both oceans and any sedimentary eras. However, according to the measured [Sr/Ba]barite and the temperature at which marine barite precipitates of both oceans, we think that the second prerequisite is not valid. Therefore we doubt that DSr provided by Averyt and Paytan [2003] is of much significance in reconstructing the Sr concentration of the ancient oceans. [2] The temperature at which marine barite was precipitated varied significantly with the geographic location of the ocean basins. The temperature of the upper water column where much of marine barite was precipitated ranges between 5 and 20°C in the equatorial Pacific and between 1 and 3°C in the Southern Ocean [Averyt and Paytan, 2003]. De Villiers [1999] and Monnin et al. [1999] also reported that the temperature range of the upper water column is quite different among the different regions of the Atlantic, Indian, Pacific and Southern Oceans. Over the interval of 20°C the percent change in DSr experimentally can attain about 15% [Church, 1979]. If assuming the temperature difference between the equatorial Pacific and the Southern Ocean at the depth of barite formation is in the range of 10°C, the expected percent change in DSr theoretically would be 10% approximately [Church, 1979; Averyt and Paytan, 2003]. The temperature can effectively affect DSr. Consequently, DSr should be obviously different among different oceans and even within the single ocean basin. [3] In addition, the variation in DSr with sedimentary era is possible. In fact, Sr/Ca-based paleotemperature estimates of corals indicate that tropical sea surface temperature (SST) during the Last Glacial was cooler up to 5°C than present temperature [Henderson, 2002]. Other proxies (i.e., alkenones and Mg/Ca) have led to a developing consensus of glacial tropics cooler by 3°C or so [Bard, 2001]. Accordingly, it is possible that there are different temperature ranges of the upper water column between the ancient and modern oceans. If so, DSr should not be a constant but varies temporally. [4] The Sr content of the barite samples fluctuates very remarkably from one region to the other and even within the same core. In the Pacific Ocean, [Sr/Ba]barite varies from 23.97 to 40.13 mmol/mol, and in the Southern Ocean the range is from 27.34 to 33.13 mmol/mol [Averyt and Paytan, 2003]. Furthermore, for the sediment core PLDS-77BX, [Sr/Ba]barite varies from 23.97 to 37.23 mmol/mol. These data together indicate that if Sr is indeed incorporated into marine barite in proportion to the value of [Sr]sw/[Ba]ss, DSr is significantly different at different regions and even at the same geographical location. Therefore it is hard to consider that Sr incorporates into marine barite in the proportion to the value of [Sr]sw/[Ba]ss if an identical DSr is available for the both oceans. [5] In light of the above facts, we think that it is not reasonable to estimate a single DSr for the world oceans and that the availability of DSr defined by Averyt and Paytan [2003] is unreliable. The spatial and temporal variation in temperature and marine chemistry should be considered seriously in reconstruction of paleoceanography. [6] This study was supported by the Key Project of Chinese Ministry of Education (grant 104089).