On the bottom of the deep sea, just a few miles from the white sandy beaches lining the Gulf of California in northern Mexico, researchers at the University of Arizona have discovered the geological equivalent to an unopened book containing untapped records about Earth's history.

Analyzing traces of leaf waxes from land plants that over millennia accumulated in deep sea sediments, a team of researchers led by the UA reconstructed the history of monsoon activity in northern Mexico. Their results, published online Sept. 3 in Nature Geoscience, help settle a long-standing debate over whether monsoon activity shut down completely or was merely suppressed under the influence of cooling brought about by the ice sheets that covered much of North America.

During the last glacial maximum, about 20,000 years ago, when mammoths and other prehistoric beasts roamed what is now northern Mexico and the Southwestern U.S., summer rains contributed about 35 percent of the annual rainfall, compared to about 70 percent today, according to the study.

By diverting moisture from the tropics, the summer monsoon brings relief from months-long intense summer heat and drought to the arid lands of the American Southwest and northwestern Mexico. If the region depended on winter rains alone, the Sonoran desert would not be known as one of the world's most biodiverse deserts.

"The monsoon is such an iconic feature of the desert Southwest, but we know very little about how it has changed over thousands and millions of years," says Tripti Bhattacharya, the study's first author. "Our finding that the Southwestern monsoon was suppressed but not completely gone under glacial conditions points to the dramatic variability of the atmospheric circulation at the time, but suggests it has been a persistent feature of our regional climate."

Previous studies had yielded inconclusive results, in part because the records used to infer evidence of past monsoon rainfall tend to be more like snapshots in time rather than providing more continuous climate records. For example, researchers have gained valuable glimpses into long-vanished plant communities based on plant parts preserved in packrat nests called middens or by analyzing the chemical signatures they left behind in soils. Those studies suggested persistent monsoon activity during the last ice age, whereas other studies based on climate modeling indicated it was temporarily absent.

By applying a clever method never before used to study the history of the monsoon, Bhattacharya and her co-authors discovered the equivalent of a forgotten, unopened book of past climate records.

Forming a vast natural vault almost 1,000 meters below the sea surface, the seafloor of oxygen-poor zones in the Gulf of California contains organic material blown into the water for many thousands of years, including debris from land plants growing in the region. Since the deposits remain largely undisturbed from scavengers or microbial activity, Tierney and her team were able to isolate leaf wax compounds from the seafloor mud.

Co-author Jessica Tierney, an associate professor in the UA's Department of Geosciences and Bhattacharya's former postdoctoral adviser, pioneered the analysis of the waxy coatings of plant leaves to reconstruct rainfall or dry spells in the past based on their chemical fingerprints, specifically different ratios of hydrogen atoms. The water in monsoon rain, according to Tierney, contains a larger proportion of a hydrogen isotope known as deuterium, or "heavy water," which has to do with its origin in the tropics. Winter rains, on the other hand, carry a different signature because they contain water with a smaller ratio of deuterium versus "regular" hydrogen.

"Plants take up whichever water they get, and because the two seasons have different ratios of hydrogen isotopes, we can relate the isotope ratios in the preserved leaf waxes to the amount of monsoon rain across the Gulf of California region," Tierney explains.

Piecing together past patterns of the Southwest monsoon can help scientists better predict future scenarios under the influence of a climate that's trending toward a warmer world, not another ice age, the researchers say.

"The past is not a perfect analog, but it acts as a natural experiment that helps us test how well we understand the variability of regional climate," says Bhattacharya, who recently accepted a position as assistant professor of earth sciences at Syracuse University. "If we understand how regional climates responded in the past, it gives us a much better shot at predicting how they will respond to climate change in the future."

One way scientists can take advantage of past climate records is by applying climate models to them, using the records to "ground-truth" the models.

"The problem is that right now, our best climate models don't agree with regard to how the monsoon will change in response to global warming," Tierney says. "Some suggest the summer precipitation will become stronger, others say it'll get weaker. By better understanding the mechanics of the phenomenon, our results can help us figure out why the models disagree and provide constraints that can translate into the future."

To test the hypothesis of whether colder times generally weaken the monsoon and warmer periods strengthen it, Tierney's group is planning to investigate how the monsoon responded to warmer periods in the past. Future research will focus on the last interglacial period about 120,000 years ago and the Pliocene Epoch, which lasted from 5.3-2.5 million years ago and was marked by greenhouse-gas levels similar to those in today's atmosphere.

Having better records of the Southwestern monsoon helps scientists better understand how it compares to monsoons in other parts of the world that are better studied.

"We now know that our monsoon appears to be much more sensitive to the large-scale configuration of the atmosphere, whereas other monsoon systems are tied more closely to local ocean conditions," Bhattacharya says.

The study, "Ice sheet modulation of deglacial North American Monsoon intensification," is co-authored by Jason Addison at the U.S. Geological Survey in Menlo Park, California, and James Murray at the School of Oceanography at the University of Washington. Support for this research comes from the David and Lucile Packard Foundation Fellowship in Science and Engineering to Tierney and NSF grant OCE-1651034 to Tierney. Addison is supported by the USGS Climate Research and Development Program.