Here's another example of how genetic methods can shed light on archaeological questions, Paleo-Eskimo mtDNA Genome Reveals Matrilineal Discontinuity in Greenland:
The Paleo-Eskimo Saqqaq and Independence I cultures, documented from archaeological remains in Northern Canada and Greenland, represent the earliest human expansion into the New World's northern extremes. However, their origin and genetic relationship to later cultures is unknown. We sequenced a mitochondrial genome from a Paleo-Eskimo human, using 3400- to 4500-year-old frozen hair excavated from an early Greenlandic Saqqaq settlement. The sample is distinct from modern Native Americans and Neo-Eskimos, falling within haplogroup D2a1, a group previously observed among modern Aleuts and Siberian Sireniki Yuit. This suggests that the earliest migrants into the New World's northern extremes derived from populations in the Bering Sea area, and were neither directly related to Native Americans nor the later Neo-Eskimos that replaced them.
New Scientist has a popular press profile of the research & findings. Remember last year when it was confirmed that Polynesians had to have been visiting the coast of South America because of the phylogeny of chicken DNA extracted from subfossils? Though there have always been hints, I think this suggests greater complexities to our picture of the pre-Columbian world. Do note that this is one mitochondrial DNA lineage. It shows lack of perfect continuity, but does not entail total replacement.... Update: "Polynesian" chickens might not be in the bag yet. See comment.