Interpretation of the seasonal variations in nitrogen isotopic signals of settling particles in the western subarctic Pacific with an ecosystem model including nitrogen isotopes

Settling particles were collected by time-series sediment trap moored at station KNOT (44N 155E, water depth 770m) during periods from October 25th 1999 to June 20th 2001, and from June 25th 2002 to May 11th 2006. Particulate nitrogen contents and its isotope ratios (d15N) in the collected samples were measured. The results of d15N showed the general variations that are lower during the spring bloom periods and higher during winter periods. In order to interpret the processes controlling such variations quantitatively, we developed an ecosystem model including nitrogen isotopes on the basis of the recent ecosystem model of Fujii et al. (Deep Sea Research, vol. 49, pp. 5441-5461, 2002), which successfully simulated the observed seasonal cycles of ecosystem dynamics at station KNOT. In our model, we took it into consideration that the isotope fractionations by nitrate and ammonium assimilation by phytoplankton, excretion and egestion by zooplankton, nitrification, remineralization of particulate and dissolved organic nitrogen (PON, DON) and decomposition of PON, which influence the d15N values of settling particles. This model was validated using an actual data set and successfully reproduced the seasonal variations in d15N of settling particles. Simulated lower d15N values in spring bloom period were mainly caused by phytoplankton assimilation of nitrate with the lowest d15N in a year. Simulated higher d15N values in winter were mainly determined by the two processes: 1) enhanced d15N values of ammonium by nitrification and following assimilation of ammonium by phytoplankton, and 2) increase of relative proportion of zooplankton, with higher d15N values compared to phytoplankton, in settling particles. Of the two processes, the former had much larger impacts on d15N values of settling particles than the latter did. In our model, we investigated the sensitivity of isotopic fractionation effect of each process to the seasonal variations in d15N values of settling particles. As a result, we found that the "°15N values of settling particles can be influenced considerably by the difference between isotopic fractionation effects of ammonium assimilation by phytoplankton and nitrification, in addition to isotopic fractionation effect associated with nitrate assimilation by phytoplankton. If the difference were found to be variable over spatial and/or temporal scales, then the variations in the difference might play an important role in the paleoceanographic record of d15N data.