Isotopic composition of particulate organic nitrogen and its relationship to nitrate assimilation in the Mediterranean Sea

The shifts in the 15N:14N ratio of particulate organic nitrogen (PON) suspended in the upper layer of the ocean have normally been interpreted in terms of changes in the isotopic composition of the source nitrogen used by the phytoplankton (nitrate, ammonium and dinitrogen gas) and/or the fractionation associated with the assimilation processes (14N is preferentially taken up by phytoplankton). The objective of the present study was to test which process, isotopic fractionation during NO3- uptake or reduced N-form consumption by phytoplankton, was more important in explaining the variations in the 15N signature of PON in the surface layer of a geostrophic front area (northwest Alboran Sea, Mediterranean Sea). Samples of PON were collected at 6 different stations during 4 seasonal surveys. The particulate organic carbon (POC) and PON concentrations averaged over the mixed layer (ML) ranged from 142 and 23 μg L-1 in summer to 104 and 13 μg L-1 in autumn respectively. The C:N molar ratio of organic matter also varied seasonally from 7.2 in summer, winter and spring to 9.3 in autumn. The range of variation of the δ15N of PON in the ML was –0.1‰ to 6.5‰. The lowest values of δ15N (lower than 1‰) were obtained in summer and winter. However, the δ15N of PON in autumn and spring was higher than 3.5‰. Consequently, the particulate matter in the ML was enriched in 15N by about 1.5‰ in autumn and spring with respect to summer and winter. The δ15N of PON in the surface layer was negatively correlated with nitrate concentration (r = -0.62, n = 22, p = 0.002). The δ15N was also negatively correlated with the proportion of nitrate to total inorganic nitrogen (i.e. the sum of nitrate, nitrite and ammonium). Furthermore, there was a negative correlation between nitrate relative to nitrate plus ammonium assimilation rates (rNO3–) and the δ15N (r = –0.71, n = 20; p < 0.001) for the surface samples. rNO3– was particularly low (lower than 0.2) for the surface samples collected during autumn, when primary productivity mainly depended on the consumption of ammonium coming from the recycling of organic matter. Therefore, δ15N of seston was a sensitive indicator of the transition from productivity based on nitrate to productivity based on regenerated ammonium.