Why does the timeline and terms need to be defined? Does it have any practical significance (other than as a topic for review)? The emergence of the written record surely did not somehow improve preservation of DNA. Applying "ancient" DNA standards is probably a good idea for "historical" specimens anyway, just to prevent contamination and get better quality material. I would imagine that a lot depends on what kind of taxa one works with, e.g. Daphnia in sediments should be preserved much better than a land invertebrate of the same age. Furthermore, if we work with human material, contamination with DNA of other humans (eg the researcher) will indeed get worse with every century, whereas for something very divergent it's probably less of an issue.

Bottom line: arbitrary terms.

I think, the history varies from population/species to population/species under comparison. If the population is preserved in ancient skeletal records, say over 500 years, but still living, it can have both ancient and extant DNA for comparison. 500 years is also a tentative minimum date I think so. Within this period I think there may not be much change if the population maintains endogamy.

There is no minimum age limit for fossils, but most paleontologists follow a rule of thumb that Pleistocene and older remains are fossil while Holocene remains are modern or subfossil. For trace fossils, this was formulated in print here:

Bertling, M., Braddy ,S.J., Bromley, R.G., Demathieu, G.R., Genise, J., Mikula´sˇ, R. , Nielsen, J. K., Nielsen, K. S. S., Rindsberg, A. K., Schlirf, M. & Uchman, A. 2006: Names for trace fossils: a uniform approach. Lethaia, vol. 39, pp. 265-286.

Dan Graur posted about this on 10/03/2015:

http://judgestarling.tumblr.com/

"Ancient DNA is defined as any genetic material recovered from biological samples that have not been preserved specifically for later DNA analyses. Thus, ancient DNA is “ageless.” One simply has to be precise about the age.

In the forensic literature, however, the term ancient DNA often refers to biological material older than 75 years (Graham 2007).

Should we be more specific and divide ancient DNA into different age groups? Some have suggested that the term ancient DNA should be reserved for things that are really, really old. Younger sources of DNA should be called antique DNA.

I am not sure adopting the language of auction houses is advisable. We may end up with post war DNA (after the end of World War I), mid-century DNA (1940s and 1950s), pop era DNA (1960s and early 1970s), modern DNA (end of 19th century to 1960s), and contemporary DNA (1980s and onward).

Graham EAM. 2007. DNA reviews: ancient DNA. Forensic Sci. Med. Pathol. 3:221-225."