Alkenone temperature records and biomarker flux at the subtropical front on the chatham rise, SW Pacific Ocean

Abstract

Alkenones and a suite of sterol biomarkers were examined in two sediment trap arrays deployed at 300 m depth in subtropical and subantarctic waters to the east of New Zealand from late winter to autumn in 1996–1997. The two traps were located within 200 km of one another and the main difference between the two sites are the differential physical, chemical, and biological characteristics of the different water masses in which they were situated. The alkenone-based reconstructions of water temperatures ( U 37 K ' ) were compared to the COADS monthly averaged satellite and real-time weekly temperatures for the deployment period. The records correlate well with seasonal sea surface temperatures (SST) for the 9 months of the deployment, with temperature reconstructions within 2 °C of regional monthly averages for most of the year. There are a few short periods of poorer agreement where alkenone-based reconstructions deviate by up to 4 °C in both traps. Weekly averages of satellite SST obtained during the time of the deployment indicate that these deviations were not associated with short-term changes in surface temperatures overlying the traps. These instances of poor correlation are not due to lateral advection of particles, but rather seem to reflect differences in environmental controls on alkenone-derived SSTs in the two water masses. Subantarctic traps showed deviations only to warmer than average temperatures. These occurred in early winter and late summer, during times of low lipid fluxes, suggesting that slow growth associated with light limitation may have affected unsaturation levels in the alkenones. The subtropical traps showed deviations only to cooler temperatures, which occurred in the late summer to early autumn. These biases occurred during times of highest lipid fluxes and lowest nutrients in the surface mixed-layer. Alkenone temperatures during maximum flux periods were too cool to be caused by subsurface production alone, suggesting that nutrient limitation effects may have significantly depressed alkenone unsaturation levels. Sterol biomarker associations indicate very different timing and trophic interactions between phytoplankton in the respective water masses. In subtropical waters, fluxes of dinosterol (a biomarker for dinoflagellates) peak first in the spring bloom while alkenones and diatom marker sterols respond synchronously during the main bloom event. In subantarctic waters sterol fluxes indicate a succession of phytoplankton export production in the spring bloom, whereas lipid fluxes normalized to organic flux indicate that alkenone producers were a relatively small proportion of the main spring bloom and proportionally more important when organic fluxes were lower in summer. We infer that the difference in nutrient concentrations between the water masses may drive these trophic differences and the inverse relationship between flux and the temperature response of U 37 K ' between the two sites.