A fire regime in the perhumid rainforest?

It's somewhat canonical that the perhumid temperate rainforest, north of about 55 degrees latitude, doesn't burn.  Wet conditions year round and low temperatures mean that even the abundant and continuous fuels don't usually catch on fire.  We just compiled a list of all known fires in southeast Alaska, and only about 10 are above 2.5 ha or so, from the late 60's to present - and all were set by people.  Generally most people assume that the forest has never really caught on fire, there's not any evidence for even a long (e.g., 500-1000 year) return interval, and most of the forest is uneven aged, unless it originated after a windstorm or landslide.

As in any field of science, it's really exciting when you find out that some assumptions are wrong.  Turns out that there is ample evidence of fires throughout southeast Alaska from about 8000-9500 years before present.  Charcoal shows up in sediments from about that timethroughout southern southeast; I'm just starting to look for it in northern southeast.  This time period corresponds to a relatively warm (comparable to strong El Nino years, like this year) and relatively dry period, the Holocene optimum.  But crucially, this appears to be after trees (Sitka spruce, shore pine, and the hemlocks) arrived, so they were forest fires, not a post-glacial max shrub or grassland system.  That's significant, as it indicates that these trees can burn, did burn, and that burning may have had some influence on relative migration rates.  Plus, it is a way into calibrating fire models for the future.  That's been a tricky problem - sure it may burn, but without known fire causing conditions in these forests, how to calibrate models to explore that possibility?  This may give us a way to do that.

Below is a snippet from a poster presented by Baichtal et al. in 2008 showing where charcoal has been found around Prince of Wales.  Stay tuned for more.

Snipped from AAA poster, 2008, Baichtal et al.  

Snipped from AAA poster, 2008, Baichtal et al.