The modern-day Namib desert has been arid for at least 15 million years, when the extinct relatives of ostriches laid their eggs, which are like time capsules of ancient air. JP Koch/Red Dune SafarisSome periods in Earth history are so different from our own that they may as well belong to another planet. Many people are interested in the age of dinosaurs, or the Ice Ages, but it is an intermediate world, the Miocene Epoch – a sort of “in-between” world, geologically speaking: less recent than mammoths and stone tools, but not the deep past of dinosaurs – that many scientists find interesting. By the start of the Miocene, the dinosaurs had been dead for 40 million years. The continents were more or less in their current positions, yet our hominin ancestors had not ventured down from the trees. It was our planet, immediately before humans – instantly recognisable, but only just. Read more: Ancient microbial life used arsenic to thrive in a world without oxygen During the Middle Miocene, around 17 million to 15 million years ago, carbon dioxide levels in the atmosphere were slightly higher than today. This was a time when greenhouse conditions prevailed: global temperatures were warmer than they are now, sea levels were higher, and water was not yet fully locked up in the great polar ice sheets. It creates an interesting picture of what global warming could be like for the planet in future. Read more: Frog fossils tell us something new about rain patterns on South Africa’s west coast One way scientists reconstruct conditions in the past is by analysing isotopes – variant forms of atoms – preserved in ancient fossils. Each isotope is found in slightly different proportions depending on the conditions in which it was locked away, making fossils a kind of molecular time capsule.Plants breathe in carbon dioxide and produce oxygen, but in doing so they preferentially take up certain isotopes, altering the proportions left in the atmosphere. Birds absorb that oxygen through breathing and eating, locking those proportions into the mineral structure of their eggshells.Together with my co-author, isotope specialist Drake Yarian, I recently extracted oxygen from eggshells laid by long-extinct giant birds, buried more than 15 million years ago in what is now the Namib Desert – the world’s oldest desert, stretching 2,000km across Angola, Namibia, and South Africa. We did this to understand how actively plants were processing CO₂ during the warmer Miocene, and what that can tell us about future climate change.Our research developed a laser-based technique that measures tiny variations in atoms of ancient oxygen directly in fossil eggshells – material far older than ice cores can reach – requiring ten times less material than previous attempts.Most predictions show that in the next few decades, Earth’s atmosphere will hold greenhouse gas levels that will meet or even exceed the levels seen in the Miocene period. Studying the Miocene therefore holds clues about what we might expect from future climate change. How plants responded to a warming climate millions of years agoAs scientists, we are interested in why carbon dioxide (CO₂) levels were higher in the past than they are now and how the plants and trees on the surface of the land responded. Scientists like us are especially interested in the rate at which plants responded to increased carbon dioxide as the planet warmed. This rate is known as primary productivity.Warming and higher CO₂ would have stimulated plant growth – and with it, the absorption of carbon from the atmosphere – but would also have accelerated the decay of plant matter, releasing carbon back. The balance between these opposing processes would have determined whether the land surface slowed down climate change or accelerated it.By the strictest definition, one would need to measure and weigh all plant matter on the planet, including in the oceans, to determine primary productivity. Even the most advanced satellites cannot detect every leaf, probe deep soils, or go back in time 15 million years. Instead, scientists rely on indirect approaches, developing stand-ins that record how active Earth’s biosphere once was.Unlocking time capsules in fossilsFossil eggshells preserve oxygen atoms in their mineral structure; a chemical record of the atmosphere as it was when the bird was alive. Our task is to liberate that oxygen and measure variations in its isotopes. What interests us is the oxygen locked inside this fossil mineral, and in particular a rare form of it known as oxygen-17. To measure it, we release it from the eggshell using acid, freeing oxygen that has been bound in solid rock for 15 million years. Read more: Africa is full of bats, but their fossils are scarce – why these rare records matter Oxygen has three stable forms: the common oxygen-16, the heavier oxygen-18, and the exceedingly rare oxygen-17. If the atoms in the air were people in a city, oxygen-16 would fill the streets, oxygen-18 a sports stadium, and oxygen-17 would struggle to fill a commuter train. Only one in roughly 4,000 atoms of oxygen is oxygen-17.Another distinctive feature of oxygen-17 is that it is readily transferred between molecules in the Earth’s upper atmosphere. There, reactions involving ozone and sunlight redistribute oxygen-17 into carbon dioxide, where it is more commonly found. Winds mix this around the planet, until plants remove the excess during photosynthesis. Read more: Plant fossils have a lot to teach us about Earth’s history During periods of intense plant growth, plants remove more carbon dioxide and ¹⁷O from the air, leaving slightly less in the oxygen we breathe and the food we eat. Changes in ¹⁷O therefore help reveal how fast the plants of that time have processed carbon dioxide.Oxygen absorbed by animals through breathing, eating and drinking becomes locked in fossils. So measuring oxygen-17 in fossil teeth and eggshells lets scientists reconstruct past primary productivity, reflecting how actively the biosphere once absorbed CO₂ and thus how much it was either buffering or amplifying the climate at that time.While ice cores offered a way to test this idea, they reach back only a few hundred thousand years, and extending the method to fossils was long hampered by the vanishingly small amounts of oxygen-17 they contain. New laser-based techniques now overcome this, detecting oxygen-17 in tiny samples faster and with far less material destroyed.What we foundWe built our prototype instrument during the first COVID-19 lockdown in 2020 and spent three years making our measurements of eggshell fossils. From dozens of measurements, we uncovered a surprising result. Around 15 million years ago, the biosphere slowed down, and plants may have been approximately 40% less active at absorbing CO₂ than today.These are initial results. The models linking oxygen-17 in fossils to the global carbon cycle are still being refined, and independent laboratories will need to replicate our findings. The pace of this kind of science is slow, because careful checking is essential.Yet the implications are urgent. Today, plants and soils absorb around a third of the carbon emissions released by human activities. As the planet continues to warm, understanding how plants respond to rising carbon dioxide becomes critical because it determines how rapidly the climate will change. Read more: How looking 250 million years into the past could save modern species It took tens of thousands of human lifetimes for Earth to cool from the warm greenhouse conditions of the Miocene into the colder world in which our ancestors evolved. In just a few generations, people have reversed that through industrial activity, and changes in how we use the land. By detecting atoms trapped in fossils, we can shed new light on Earth’s past, and perhaps better understand the future now unfolding before us.Vincent Hare receives funding from the National Research Foundation of South Africa, the Biogeochemistry Research Infrastructure Platform (BIOGRIP) of South Africa, and additional support from the International Atomic Energy Agency.