Nitrogen fixation in the tropical Atlantic weakened during the warm late Pliocene, fossil shells show

The proxy is a trace of nitrogen locked into shells. Foraminifera take up small amounts of it as they grow their calcareous skeletons, and those skeletons survive in seafloor sediment, so a nitrogen signal can be read off animals that died millions of years ago ^[3]. Earlier work from the same institute had reconstructed the other half of the cycle, showing that water-column denitrification weakened globally in past warm climates and leaving the fixation side open ^[7].

That gap is where the phosphorus link matters. Denitrifying bacteria strip nitrate out of oxygen-poor water and leave the residue comparatively rich in phosphorus. Currents redistribute that surplus, and nitrogen fixers in nutrient-poor surface water draw on it, so a slower Pacific sink means less phosphorus arriving in the Atlantic ^[9]. The support offered for that chain is that the Atlantic fixation record and a recently published Pacific denitrification record both fall in the late Pliocene ^[8]. Two reconstructions declining across the same interval fit the phosphorus explanation, but on their own they do not separate it from other things that changed in the late Pliocene ocean.

Nitrogen fixers work mainly in warm, nutrient-poor surface water ^[4]. A warmer ocean has more of that water, and the expectation of more fixation follows easily. The tropical Atlantic record for the late Pliocene, an epoch globally warmer than today, goes the other way ^[1][13]. The phys.org account does not give the size of the decline, only its direction and timing ^[20].

This record does not resolve how quickly fixation responds, or what a nutrient-poor gyre does over a few decades. After large Northern Hemisphere ice sheets began to grow about 2.8 million years ago, Caribbean fixation rose and fell in step with the roughly 41,000-year cycle in Earth's axial tilt ^[15]. Half that cycle is about 20,500 years ^[18]. The pacing shows up roughly 220,000 years before the Pliocene ends ^[17], inside an epoch that ran about 2.75 million years ^[16].

Maayan Yehudai, a postdoctoral researcher at the Max Planck Institute for Chemistry and the study's first author ^[12], framed the projection in conditional terms. "In a warmer future, we suspect both processes will slow down, so the ocean would fix less nitrogen but would also lose less," said Yehudai, adding that "the amount of nutrients supplied to a given region may differ from today" ^[11].