Oceanographic considerations for the application of the alkenone-based paleotemperature U 37K' index in the Gulf of California
Alkenone concentrations were determined in sediment trap samples and surficial sediments collected from Guaymas Basin (Gulf of California) to investigate the factors that control the relationship between the U 37K' index and SST. The results from the sediment trap study reveal a positive correlation ( r2 = 0.5) between alkenone fluxes and satellite AVHRR sea surface temperature (SST) and a strong correlation ( r2 = 0.9) between U 37K' and SST over an annual temperature range of 20°C. Although SST and U 37K' are tightly coupled throughout two upwelling seasons, the relationship deviates from the expression developed by Prahl et al. (1988) at temperatures higher than 26°C. The possible reasons for this deviation are varied but include the observed changes in the coccolithophore species assemblage and the evolution of the thermocline during periods of high SST. The seasonal pattern of alkenone fluxes (higher fluxes during warm periods) observed throughout the 1996 to 1997 study period causes a "bias" in the annually averaged U 37K' record, resulting in temperature estimates that are ∼1°C higher than the annual AVHRR SST average. Furthermore, the U 37K' ratios obtained from core-top sediments indicate an additional ∼2°C "warm bias" in the sedimentary record that may be related to the lateral inputs of alkenones from other regions of the Gulf. These results highlight the importance of oceanographic processes in determining the alkenone temperature signal that is preserved in sediments. After accounting for such effects, U 37K' ratios from a box core were used to reconstruct a SST record extending back to the 18th century for the Guaymas Basin. These results indicate a steady 2°C warming of the surface waters of the Gulf from early 1700s to the 1950s, followed by a rapid apparent 1°C cooling between the 1950s and the 1980s.