A hydraulic monitor thaws permafrost with high-pressure water in the Klondike goldfields, Yukon — mining exposures that researchers used to access ancient ground ice and fossil records. (Duane Froese), CC BY-NC-NDPermafrost usually hits the news as a hazard; a planetary risk. When this ice-rich ground thaws, it damages roads and building foundations. It drains lakes and tips trees into drunken forests. It releases greenhouse gases that have been locked in carbon-rich soils for thousands of years, amplifying warming, driving further thaw. These are serious concerns, and we spend a good deal of time studying their causes. But another consequence of permafrost thaw receives far less attention: As the climate warms, we are losing our frozen heritage.Permafrost contains the most exceptional archive of the distant past on the planet, a record of ancient life preserved in a way that exists almost nowhere else on Earth.A recent study we led, published in Nature Communications, recovered DNA from Arctic ground squirrel droppings preserved in permafrost in Klondike, Yukon. In these pellets, we found plants the squirrels had eaten, parasites they carried and even traces of the larger animals — mammoths, horses and bison — that shared their grassland. A single dropping held a detailed portrait of a vanished ecosystem, spanning an extraordinary depth of time — from around 14,000 to 700,000 years ago.Fossils in the ‘muck’In the Klondike region of central Yukon, within the traditional territories of the Tr'ondëk Hwëch'in and the First Nation of Na-Cho Nyäk Dun, placer gold miners commonly use high-pressure water to strip away frozen silt and organic material to reach the gold-bearing gravels beneath. As the permafrost thaws, it gives up what it’s held for tens of thousands of years: wedges of ground ice; mats of grass roots still woven together; the nest of an Arctic ground squirrel with the animal’s bones curled inside. Every so often, we find the mummified carcass of an ancient wolf pup or the blue-stained glint of a mammoth tusk or a bison skull. An Arctic ground squirrel nest with a grassy lining and the partial skull of Arctic ground squirrel. (Duane Froese), CC BY-NC-ND Miners call this frozen overburden “muck.” Scientists refer to it as Yedoma — part of an extensive complex of fine-grained frozen sediment stretching from Yukon across Alaska and the Bering land bridge into Siberia. It is one of the richest paleontological and paleoenvironmental archives in the world — and we are only beginning to grasp how much it holds. A molecular archive of lifeFor most of its history, palaeontology was a hands–on science: a fossil was something you could pick up — a bone, a tooth, a tusk. Over the past decade, that’s changed. Today we examine these deposits at the molecular level. From a teaspoon of frozen soil or the droppings of long-dead animals, we can recover billions of fragments of ancient DNA directly. Permafrost preserves not just bone but the fragile, molecular archive of life — DNA, RNA, proteins, a genetic record of entire communities — under conditions of cold, darkness and low oxygen that stop the decay that would otherwise erase this material. An Arctic ground squirrel skull thawing from permafrost. (Duane Froese), CC BY-NC-ND The DNA and proteins are heavily fragmented, but new sequencing technologies and enrichment methods now allow us to reconstruct full genomes and whole communities of ancient organisms from these scattered pieces — a feat impossible even a few years ago and one that is advancing quickly.Discovery of antibiotic-resistant genesEcosystems are only one part of what is held there. Years ago, some of us were part of a team, led by researchers Hendrik Poinar and Gerry Wright at McMaster University, that searched this same ground for antibiotic-resistant genes. In 30,000-year-old permafrost, we found a diverse set, including genes associated with resistance to drugs such as vancomycin, with the ancient versions functioning much like those in hospitals today. That work showed antibiotic resistance is far older than the clinical use of antibiotics. It also hinted that this frozen ground is a vast, almost entirely unread catalogue of microbial chemistry — one that may yet inform how we counter drug resistance or yield useful new enzymes.Thaw is accelerating as ground warms across the North. As permafrost temperatures climb, microbial activity increases and more unfrozen water becomes available, allowing organic material to begin decomposing even before complete thaw. Material sealed away for thousands of years meets oxygen, free water and microbes, and the record fragments. Read more: Canada’s North is warming from the ground up, and our infrastructure isn’t ready Archives of frozen sedimentPractically, this means we need to be collecting and preserving far more of this material than we do now — building archives of frozen sediment and the molecules within them while they can still be read. A hydraulic monitor washes away frozen ‘muck’ from a permafrost exposure in the Klondike goldfields, central Yukon, revealing sediments that have preserved ice age life for tens of thousands of years. (Duane Froese), CC BY-NC-ND We have made efforts toward this goal in the Permafrost Archives Laboratory at the University of Alberta, one of the few facilities of its kind in the world, but it remains a small collection against a vast Arctic. The science to interpret these archives is advancing quickly but the deposits themselves are degrading, particularly in areas of warm permafrost such as central Yukon.The only real solution to permafrost thaw is to reduce greenhouse gas emissions and slow the warming of the North. In the absence of that, the least we can do is preserve what we can of this archive — in cold storage and in careful study — before significant portions are degraded beyond recovery. The invisible DNA archives of a lost world remain buried in these frozen soils. If those records thaw, the beauty and complexity of life that shifted and adapted to the great ice ages of the past, with lessons for our warming world, will be lost.Duane Froese receives funding from the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Canadian Foundation for Innovation (CFI). Hendrik Poinar receives funding from NSERC, SSHRC, McMaster University, IIDR.Scott L. Cocker receives funding from the Knut and Alice Wallenberg Foundation and the Swedish Research Council. Tyler J. Murchie receives funding from the Tula Foundation through the Hakai Institute and the Natural Sciences and Engineering Research Council of Canada (NSERC).