It is commonly believed that trees were absent in Scandinavia during the last glaciation and first recolonized the Scandinavian Peninsula with the retreat of its ice sheet some 9000 years ago. Here, we show the presence of a rare mitochondrial DNA haplotype of spruce that appears unique to Scandinavia and with its highest frequency to the west-an area believed to sustain ice-free refugia during most of the last ice age. We further show the survival of DNA from this haplotype in lake sediments and pollen of Trøndelag in central Norway dating back ~10,300 years and chloroplast DNA of pine and spruce in lake sediments adjacent to the ice-free Andøya refugium in northwestern Norway as early as ~22,000 and 17,700 years ago, respectively. Our findings imply that conifer trees survived in ice-free refugia of Scandinavia during the last glaciation, challenging current views on survival and spread of trees as a response to climate changes.

Birks et al. question our proposition that trees survived the Last Glacial Maximum (LGM) in Northern Scandinavia. We dispute their interpretation of our modern genetic data but agree that more work is required. Our field and laboratory procedures were robust; contamination is an unlikely explanation of our results. Their description of Endletvatn as ice-covered and inundated during the LGM is inconsistent with recent geological literature. W e thank Birks et al.(1) for their comments on our paper (2) and continuing engagement in the long-standing debate on tree survival (or otherwise) in glacial-age Scandinavia (3). We know as well as any that the new field of environmental ancient DNA requires further work; however, we find their specific criticisms to be unfounded. For our results to be merely a set of errors, as might be construed from their Comment, would require contamination by independent spruce and pine genetic markers to occur in two independently obtained sediment samples, including pollen isolates, and/or in three independent ancient DNA laboratories. Additionally, for our interpretation of the genetic patterns to be erroneous, a highly complex alternative explanation of modern spruce genetic data in northern Europe must be invoked.(i) How reliable is ancient DNA (aDNA) from lake sediments? It is a misinterpretation of Gilbert et al. (4) that a specific set of authentication criteria is needed. They recommend adding criteria of relevance to a given study rather than following a predefined list. We (i) took material from two locations with different equipment on different expeditions (minimizing cross contamination); (ii) used two independent DNA markerschloroplast DNA (cpDNA) and mitochondrial DNA (mtDNA)-for each locality (minimizing the chance that positive findings of conifer DNA result from contamination of amplicons obtained from the other core); and (iii) undertook replication of findings in three ancient DNA laboratories (two in Uppsala and one in Copenhagen).No piston corer is contamination-free, including that of (5). We followed all conventional protocols for avoiding contamination: careful core storage, cleaning of cores before sampling, and sampling from the undisturbed interior of a core in clean laboratories and with clean instruments (5). The lacustrine Endletvatn sediments were retrieved from a mire, so the corer was never in a lake water column; dissolved DNA from lake water is not a contaminant. Were contamination present, taxa currently abundant in the catchment (e.g., Calluna vulgaris and Vaccinium spp.) should have been found, but they were not; rather, most taxa were similar to pollen taxa.We do not report the plant fragment retrieved from lake sediment as contamination. It was bleached to remove modern contamination (standard procedure before ancient macrofossil genotyping) and identified by DNA barcoding as Urtica.(ii) Was glacial survival on Andøya possible? Our record is physically feasible. Current understanding of Last Glacial Maximum (LGM) glacial limits places northern An...

Owing to proximity of the North Atlantic Stream and the shelf, the Andya biota are assumed to have responded rapidly to climatic changes taking place after the Weichselian glaciation. Palynological, macrofossil, loss-on-ignition, tephra and 14 C data from three sites at the northern part of the island of Andya were studied. The period 12 300-11 950 cal. yr BP was characterized by polar desert vegetation, and 11 950-11 050 cal. yr BP by a moisturedemanding predominantly low-arctic Oxyria vegetation. During the period 11 050-10 650 cal. yr BP, there was a climatic amelioration towards a sub-arctic climate and heaths dominated by Empetrum. After 10 650 cal. yr BP the Oxyria vegetation disappeared. As early as about 10 800 cal. yr BP the bryozoan Cristatella mucedo indicated a climate sufficient for Betula woodland. However, tree birch did not establish until 10 420-10 250 cal. yr BP, indicating a time-lag for the formation of Betula ecotypes adapted to the oceanic climate of Andya. From about 10 150 to 9400 cal. yr BP the summers were dry and warm. There was a change towards moister, though comparatively warm, climatic conditions about 9400 cal. yr BP. The present data are compared with evidence from marine sediments and the deglaciation history in the region. It is suggested that during most of the period 11 500-10 250 cal. yr BP a similar situation as in present southern Greenland existed, with birch woodland in the inner fjords near the ice sheet and low-arctic heath vegetation along the outer coast.

No abstract

No abstract