Finally, changing the ratio at which ligands are released with respect to carbon (Fig. 6c) leads to a more uniform change in the average ligand profile. These characteristic

sensitivities of the ligand profile lead to corresponding sensitivities of the iron distributions. selleck kinase inhibitor Fig. 7 shows the covariation of globally averaged ligand and iron concentrations for the sensitivity runs. On the left we show the average over the whole water column, on the right the average over the top 50 m. The left plot in Fig. 7 shows that — independent of which parameter we change — the change in total iron content in the ocean is tightly coupled to the change in total ligand content, with all sensitivity experiments falling nearly on one line. It is interesting to note that in the global average, iron concentrations fall below the 1:1 line, i.e. Selleckchem BTK inhibitor the ligand excess L⁎ is always positive. A similar linear relation between dissolved iron and ligands has been also found in in-situ data from the Bering Sea ( Buck and Bruland, 2007). Of all the sensitivity experiment, changing the photochemical degradation rate (by a factor of 5) has the least

effect on global ligand and iron concentrations, which is mostly because changes are limited to the upper ocean only. Changes in average ligand and iron concentration near the surface (right plot in Fig. 7) are less universally coupled: While an increase in ligand always leads to an increase in iron and vice versa, the slopes of the relations are significantly different. A decrease in ligands through an increase in photochemical degradation affects ligand concentrations most strongly in the subtropical Pacific, with high mixed-layer irradiances and low production. Here iron concentrations are low anyway and decreasing ligands does not lead to further decreases. Paclitaxel Decreasing ligand to carbon ratios, on the other hand affects ligand production everywhere, also in regions where they affect iron residence time strongly, and hence lead to a stronger iron reduction. The number of open-ocean observations of iron-binding ligands has steadily increased

over the last decade or so, and will further do so as the international GEOTRACES program continues. One clear result of these in-situ measurements is that iron-binding ligands show substantial spatial variability in ligand concentrations between different oceanic regions (1 to 10 nM, Gledhill and Buck (2012)). In contrast, ocean biogeochemical models mostly still assume a uniform and comparatively low ligand concentration (typically between 0.6 and 1 nM). There are some exceptions to this (Tagliabue and Völker, 2011 and Misumi et al., 2013), but even these newer studies rely on empirical relationships and do not attempt to describe the sources and sinks of ligands prognostically, despite the existence of a conceptual model for their dynamics (Hunter and Boyd, 2007 and Ye et al., 2009).