This is a question that has long puzzled researchers working in Antarctica. Ice-free areas of most of Antarctica are so barren and dry under present conditions, it is hard to imagine that a tundra environment once thrived. Plants and ecosystems that characterize relatively warmer sub-Antarctic regions today, flourished across regions in the Transantarctic Mountains and down to the coast through much of the Early and Middle Miocene (Fig. 1). During this time, interglacial periods were warm and wet enough to support liverworts, mosses, and shrubby trees including Nothofagus (southern beech), which is now extinct in Antarctica. Conversely, under modern conditions, the high elevations of the McMurdo Dry Valleys are amongst the most inhospitable environments on Earth. Mean annual air temperatures hover around -20°C and ground temperatures remain below 0°C year-round, so that the ground remains permanently frozen and liquid water is absent.
The complex puzzle of Antarctic climate evolution
Reconstructions of past climate from geological data across the world show our planet experienced a period of peak warmth between 17 and 15 million years ago. This “Miocene Climate Optimum” was followed by the Middle Miocene Climate Transition, during which global climate started cooling (Fig. 1). In Antarctica, this period was characterized by increasingly extensive episodes of ice sheet advance during glacial periods and the ultimate disappearance of terrestrial plants at high elevation sites approximately 14 million years ago. Many studies also suggest that it was during this cooling period that high elevation sites in the McMurdo Dry Valleys transitioned to their current hyper-arid state. But clear and compelling evidence for the onset and duration of aridification remained elusive. When did the high elevations of the McMurdo Dry Valleys become arid and how long have they remained permanently frozen? This is the question I wanted to address when I joined the Friis Hills Drilling Project.
My piece to the puzzle
I had just finished my M.Sc. in permafrost geochemistry at the University of Ottawa in Canada, when my supervisor, Professor Denis Lacelle, asked me if I would be interested in joining a team of kiwi researchers investigating the oldest permafrost on Earth. He had been contacted by Dr Warren Dickinson, a geochemist who had conducted pioneering work to address the permanent aridity question in Antarctica and who became my PhD supervisor after I gladly accepted the offer. In 2017, I moved across the world and joined the Antarctic Research Centre in Wellington, New Zealand, to embark on my PhD journey. I had been working in the Canadian Subarctic, where the permafrost is dynamic and highly responsive to changes in climate. Thinking that an environment could have remained frozen for millions of years really had me scratching my head.
Led by Professors Richard Levy and Timothy Naish, the Friis Hills Drilling Project had just completed its drilling phase during the 2016-2017 austral summer (Fig. 2). The main objective of the project was to recover a continuous record through a unique terrestrial sequence of glacial-fluvial sediments spanning the end of the Miocene Climatic Optimum and Middle Miocene Climate Transition. Although offshore Antarctic marine records and deep ocean sequences through both periods had been well studied, terrestrial records from Antarctica are incredibly sparse. Fortunately, a suitable site, the Friis Hills was discovered in the 1990s, and then painstakingly described, mapped, and sampled by Dr Adam Lewis and Professor Alan Ashworth who pieced together a remarkable sequence from shallow surface pits and sporadic outcrops. The potential to uncover hidden parts of the sequence are ultimately what led to the Friis Hills Drilling Project, several season later.
While the primary aim was to uncover a detailed history of environmental changes from the mid-Miocene, the permanently frozen surface beneath which the team drilled also offered potential to capture chemical traces of climatic change that have occurred since the sediments were deposited. High elevation surfaces across the McMurdo Dry Valley region have been exposed since the mid-Miocene, making them suitable for cosmogenic dating. Meteoric Beryllium-10 is formed in the upper atmosphere and delivered to the surface of the Earth through precipitation. It then adsorbs onto clay particles and can only be displaced by liquid water. By looking at meteoric Berrylium-10 concentrations in the subsurface, we could then date the last time water infiltrated below the surface. In other words, we would use meteoric Beryllium-10 as a tracer for water infiltration. Under the guidance of my second PhD supervisor, Dr Kevin Norton, I spent months in the lab leaching meteoric Beryllium-10 from sediment samples from the Friis Hills and other high elevation sites in the McMurdo Dry Valleys. I packed tiny amounts of Beryllium into tiny targets for accelerator mass spectrometry dating. In November 2019, unaware of the imminent global pandemic, I took my samples to the Laboratory of Ion Beam Physics at ETH Zurich to work with state-of-the-art equipment in collaboration with Dr Marcus Christl and his team. My samples were given royal treatment as they were the first to go through their new Multi-Isotope-Low-Energy-System (MILEA).
Borders shut in New Zealand a few months after my visit to Switzerland and I had to make sense of the results I had received from Marcus. Through our study, we had analyzed 64 samples collected from 10 different boreholes at three locations in the McMurdo Dry Valleys, some dating back to 1997, using two chemical protocols and two different accelerator mass spectrometry laboratories. Science in Antarctica is all about collaboration, and as such, our finding is an added piece to the complex puzzle about climate evolution in Antarctica. It was clear from the meteoric Beryllium-10 concentrations that our sites had not remained frozen since the mid-Miocene. When were the McMurdo Dry Valleys of Antarctica last wet? Well, our study suggests that even at high elevations, water infiltrated the ground until the late Miocene, much later than previously suggested. And why should we care? Determining when this landscape became arid is critical to our understanding of the response of glacial systems in Antarctica to changes in surface temperatures. In short, this finding implies that the McMurdo Dry Valley are not a landscape frozen in time and that they are more susceptible to climate change than previously anticipated.