The Southern Ocean is iron-limited. Phytoplankton there have everything else — light, nitrate, phosphate — but not enough iron to grow. The long-standing expectation was that melting Antarctic ice would fertilize this ocean: icebergs carry iron-rich sediments, iron feeds algae, algae absorb COâ‚‚, carbon sinks, warming slows. A planetary thermostat. Negative feedback. The ice sheet's own destruction would partially counteract the warming that caused it.
A 2026 study published in Nature Geoscience examined sediment cores spanning multiple glacial cycles and found the opposite. When the West Antarctic Ice Sheet retreated during past warm periods, more icebergs calved and drifted north, delivering more sediment to the Southern Ocean. But the iron in those sediments was highly weathered — chemically altered over time by exposure to air and water in the glacial bed. Weathered iron is poorly soluble. Phytoplankton cannot use it. More delivery, less biology. The correlation between ice retreat and algae productivity was negative, not positive.
The implication reverses the climate story. Continued West Antarctic ice loss from current warming will erode increasingly weathered rock layers, delivering increasingly unusable iron. Carbon uptake in the Pacific sector of the Southern Ocean could decrease relative to today. The expected negative feedback — ice loss buffering its own cause — is actually a positive feedback. The thermostat is wired backwards.
The general principle: a resource that is present but chemically inaccessible is not a resource. The quantity of a nutrient, a signal, or a raw material tells you about potential supply. Whether the system can actually use it depends on form, not amount. When models assume that delivery implies availability, they can predict the wrong sign for the feedback — turning a brake into an accelerator, a stabilizer into a destabilizer. The distinction between having and being able to use is not a detail. It is the mechanism.