I'm really excited about this one.
Remote sensing allows us to make many observations across broad areas. This has been immeasurably useful in looking at characteristics of disturbances, burn severity, enumerating disturbance extent, etc. Most of the time, though, these are analyses of single or a few events, and disturbance ecology still suffered from the small sample size curse. It was hard to speculate about trends because each event was unique in time and space. With the publication of Hansen et al's brilliant forest change dataset, though, we have the ability to really analyzing disturbances statistically, by looking at many thousands of events over broad spatial scales.
Southeast Alaska is ideal for doing just that- a gap phase disturbance regime, broad area with similar forest types (Sitka spruce - western hemlock, the mix that goes all the way through Oregon's coast), and broadly similar climate mean we can look across the region and basically assume that ecologically processes are broadly similar. By then looking at many thousands of disturbance events (possible because of the high resolution of the Hansen et al. dataset), we can analyze disturbance [statistical] distributions. This wouldn't be possible in a place where the average disturbance size vs. landscape area was overly large, because the distribution would be skewed to those events - we need to pick a good area, and this is a good area.
This paper looks at how disturbances and forest recovery are distributed and checks it against the assumption that if things were more or less unchanging, those distributions (again, across thousands of events) will be broadly similar. Turns out in the southern portions of the region, they are. That's a nice check on our assumption, and an interesting result in and of itself (note we are excluding anthropogenic disturbance).
In the northerly portions, though, there are substantial differences which indicate an expanding forest, especially at higher treeline elevations and low elevations associated with glacial recession. We checked this spatial pattern against USFS Forest Inventory data, and found broad correspondences - the forest is expanding and putting on biomass, about 1 teragram per year. Not only that, it's expanding in some areas and contracting in others - small shifts but when integrated across such a large region, things add up. This is how forests will change (for the most part), through these imbalances in spatio-topographic disturbance and growth patterning. This is a fun line of analysis which hasn't really been explored anywhere, and I hope to continue this type of investigation. It could really spread into other areas, keeping in mind limitations associated with disturbance size vs. landscape size and disturbance size vs. frequency of course.
Fun, fun stuff! You can see a figure from the paper a few posts back. Email for more info.
Buma B, Barrett T. 2015. Signs of disturbance disequilibrium and directional change in the world’s largest temperate rainforest. Global Change Biology. In press.