All experiments used the fertilization technique employed in IronEx: releasing a solution of weakly acidified FeSO4 into the propeller wash of the ship steaming along tracks at intervals of 1-3 km, for instance, spiralling outward from a drifting, surface-tethered buoy. One would expect the weakly acidified iron solution to be quickly neutralized by mixing with alkaline seawater and the dissolved iron oxidized to its insoluble state, i.e. colloidal rust particles that would be difficult for phytoplankton to take up, but results from the experiments demonstrate otherwise. For example, the C : Fe uptake ratio exceeded 15 000: 1 in the SEEDSI experiment (Boyd et al. 2007), similar to values for natural blooms (Blain et al. 2007) and laboratory experiments, demonstrating that the physiological ability to use iron is present in the phytoplankton. In other words, we see no need yet to replace FeSO4.
Small-scale experiments employing several tonnes of FeSO4 have not lost their relevance because they can be carried out at regional scales to test whether the plankton of a given water body in a given season is iron limited or not. For instance, such experiments will help in elucidating the causes of the regime shift that has occurred in the eastern Bering Sea accompanying retreat of the winter ice cover (Smetacek & Nicol 2005). Nutrient and light availability have not changed, yet the annual phytoplankton bloom, which used to occur during melting of the ice cover in early spring, is now delayed by several months resulting in a shift in the timing of the annual peak of food supply to the benthos.
Further testing of the iron hypothesis will require larger-scale experiments at sites where the fertilized surface layer is coherent with the underlying deep-water column. Such conditions are met within the closed cores of mesoscale eddies that are formed by meanders of frontal jets that enclose a water mass from the adjacent branch of the ACC. Eddies are of two types depending on the direction of rotation: clockwise-rotating eddies enclose a water column - the core of the eddy - from south of the respective front. Owing to its lower temperature compared with the surrounding frontal jet, the core has a smaller volume and appears as a depression in altimeter images of sea surface height. The opposite holds for anticlockwise eddies that appear as bulges in the images. Such quasi-stationary eddies extend to the sea floor and can have lifetimes of several months. They are approximately 100 km in diameter (including the enclosing frontal jet) and 4 km deep, and hence have to be visualized as slowly rotating, flat discs completing a revolution once a week. Such mesoscale eddies are ideal for studying the growth and demise of iron-fertilized blooms all the way down to the underlying sediments but their lifetime is too short for longer-term experiments over the growth season.
Experiments carried out in closed cores of eddies have the added advantage that the fertilized patch retains its initial circular shape over the course of the experiment (Cisewski et al. 2008) and is not distorted into a streak as happened during SOIREE and the SOFEX north patch (Boyd et al. 2000; Bishop et al. 2004). Furthermore, the closed core tends to be horizontally homogeneous in its properties; hence, the effects of local patchiness are greatly diminished. The disadvantage is that horizontal mixing within the core results in the patch spreading rapidly, hence diluting the effects of fertilization by mixing with non-fertilized water from within the eddy core (Cisewski et al. 2005). The dilution effect can be lessened by taking care to position the patch as close to the centre of the eddy as possible, where current speeds are lowest, and hence coherence between the surface and the underlying deep-water column greatest. Unfortunately, the connection with the sediment surface is looser because eddies can be displaced laterally by approximately 30 km in their lifetime (Losch et al. 2006). The dilution effect can also be lessened by fertilizing as large an area as feasible from a research ship. The larger the area of a patch the lesser the effects of dilution at its centre. However, the area of the closed core of an average eddy of 100 km diameter is approximately 3000 km2, which is approximately 5-50 times the area of the patches fertilized in previous experiments. Clearly, fertilizing the centre of the core is the best option but locating it at sea is not a trivial task.
Was this article helpful?