North Pacific Paleotemperature and Paleoproductivity Reconstructions Based on Diatom Species

Abstract

AbstractRegionally developed diatom‐based transfer functions reconstruct both summer sea surface temperature and primary productivity (PP) in the North Pacific based on a calibration data set including the eastern subtropical gyre (California and Oregon), the Gulf of Alaska, the western Pacific, and Bering Sea. Factors of the floral assemblage and environment data sets, and canonical correspondence, support estimation of two essentially orthogonal variables. Validation of the summer sea surface temperature transfer function resulted in r2jack of 0.9 and a root‐mean‐square error of prediction of 1.0°C over a range of 7–17°C. The summer PP transfer function yielded r2jack of 0.7 and a root‐mean‐square error of prediction of 219 g C·m2·year over a range of 200–2,000 g C·m2·year. Application in downcore locations within the modern nearshore upwelling regime and within the California Current offshore Oregon reveals that during Holocene time higher PP (>800 g C·m2·year) was associated with cooler temperatures (<16°C) forced by upwelling. In contrast, during the Last Glacial Maximum, when subpolar conditions are present off Oregon, higher PP (>1,000 g C·m2·year) was associated with warmer SSTs (>10°C; significant in the nearshore site). This finding supports previous hypotheses that in cold regimes warming may augment biological production, perhaps by relieving metabolic constraints. This is of particular interest because of projected amplification of future warming in high latitudes and the potential for greening of the subpolar and polar oceans.