Erin McClymont’s laboratory has six kitchen freezers, each alarmed. This is for self-preservation: when the power goes out or an appliance breaks down, she needs to act quickly to prevent the nauseating smell of old fish from seeping out. The source of that stench: solidified blocks of 50,000-year-old regurgitated stomach oil from Antarctic Snow Petrels, hunks of which line each freezer’s shelves.

“One of my colleagues who did some of the original sampling..., he’s been on field work where they’ve had to dump their coats at the end because they can’t get the smell out,” says McClymont, a paleoclimatologist at Durham University in England. The blocks “are revolting.”

The solidified oil is an indirect record, or proxy, of the past that scientists rely on “because we don’t have a time machine,” says Tyler Karp, a paleoecologist at the University of Chicago. Researchers trying to understand Earth’s climate and ecosystems need to trace rainfall, ice coverage, fire and other factors over thousands or millions of years—far longer than human records. But the most common proxies, including tree rings, pollen and ice cores containing pockets of ancient air, have already been well studied. To learn something new, researchers have to get creative.

On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.

So scientists like McClymont look for clever ways to study obscure features of the past, such as her research focus of how the Antarctic environment has changed over tens of thousands of years. Sea ice makes drilling samples from the Antarctic seafloor particularly hard to do, so her collaborators found a more obscure polar record to explore: seabird stomach oil. Snow Petrels spit out the oil in front of their nests, primarily to ward off predators with its smell and stickiness. That vomit accretes in layers across avian generations, trapping 50,000 years’ worth of data about the birds’ diet and the sea ice environment.

Seabird oil is a good proxy because it’s composed of waxes and fats, which degrade more slowly than proteins and carbohydrates do. Chunks can be radiocarbon-dated and biopsied to determine the source of their carbon and nitrogen—a process that first requires using a large saw to cut through what feels like “a mild cheddar: slightly soft and squishy,” McClymont says. As sea ice shrinks and expands, the surface ocean warms or cools. That in turn affects how nutrients cycle and where different species can live, which shows up in the birds’ regurgitations.

Such methods have revealed that as Antarctic ice sheets expanded during the last glacial maximum, sea ice got pushed farther offshore, forcing krill to move out of the petrels’ feeding ranges. The finding suggests that Snow Petrels are capable of temporarily adapting to different food sources and might do so again during future climate changes.

Tripti Bhattacharya, a paleoclimatologist at Syracuse University, also takes advantage of waxy substances—though much less smelly ones. She puts fossilized, freeze-dried sediments through a “glorified espresso machine” to extract the hydrophobic outer coating of ancient leaves, called leaf wax. As plants use rainwater to grow, their leaves pick up the rain’s characteristic ratio of hydrogen isotopes—atoms with the same number of protons and different numbers of neutrons. Hydrogen isotope ratios, specifically, can be traced back to reveal how much and when water fell onto a plant.

Bhattacharya found that the last time carbon dioxide levels in the atmosphere went above 400 parts per million, as present-day levels do, southern California had rainy summers; today it has rainy winters. The finding helps explain the hot, humid environment that made California suitable for tropical animals such as crocodiles three million years ago. Climate change likely won’t bring crocs back to the West Coast, but ecologists can use the ancient weather data to forecast the kinds of species that might thrive in a similar future. “Proxy data might seem like this obscure scientific thing, but it actually directly helps our efforts to manage climate risk,” Bhattacharya says.

Rainfall inspires even stranger proxies than leaf sediments: when ostriches eat plants that grew in rainy conditions, a signature isotope ratio of nitrogen in the soil transfers into their body and, ultimately, their eggs.

Princeton University geochemistry Ph.D. student Mingzhe (Damon) Dai collaborates with archaeologists to obtain samples of ostriches’ eggs buried among early human settlements whose inhabitants ate the birds and used the eggshells as water containers. The eggs’ nitrogen isotope ratios can thereby help reconstruct the rainfall experienced by early humans across Africa and Asia.

Dai’s initial results have revealed that rainfall was low in South Africa during the last glacial maximum and increased as the planet warmed. Dai says that changes in human culture and behavior occurred at the same time as these climate shifts, suggesting that they could have been an important driver of such alterations to the way people live.

Although scientists can use conventional records such as sediment and seafloor cores to reconstruct the global climate during civilization’s early days, “the resolution is too [coarse], so it’s not that fun for us to discuss something that happened on the local scale, which is relevant to the human story,” Dai explains. Coincidentally, ostriches and early humans moved within roughly the same area in their lifetime: 85 square kilometers. If an ostrich egg reveals rainy conditions, that means an individual human settlement likely also felt that precipitation—and maybe the inhabitants changed their behavior accordingly.

Karp likes how these unusual records enable him to “use the past as an experiment.” He primarily studies chemical remnants of burned plants, called polycyclic aromatic hydrocarbons (PAHs), to trace ancient fires. He also looks at chemicals called stanols, which are found in herbivore dung and help reveal when animals such as elephants, zebras, hippopotamuses and impalas roamed the land. As ecologists contemplate reintroducing fire or animals to landscapes where human activity has removed them, Karp says it’s helpful to check such strange records to see how environments responded in the past, like in a natural field trial.

Ultimately, every proxy has its limitations; some, such as the PAHs, can come from multiple sources, so it’s hard to ensure that you’re tracking the right phenomenon. Others, such as the seabird oil, may be older than radiocarbon dating can divulge. And in general, you need several detailed experimental steps to link such an unusual record to the pattern you’re trying to study, Bhattacharya says, which adds uncertainty and error at each step.

But for many researchers, these ancient remnants are often the only clues available. “I don’t think when I first started studying science [that] I would have expected that I’d be studying poop,” Karp says. But “the more different tools you can use to look at the same question from more angles, the better and better we get toward having a good consensus on what actually happened.”