Good. It's an unclear term.
This has nothing at all to do with which or how many trees are felled, but instead how they are "yarded" out of the forest. For example, helicopter yarding belies that assertion as does "high lead" cable yarding where the logs are carried in the air from a suspended cable. A rubber-tired loader moves less dirt than a crawler. Both can be used to improve conditions on the forest floor. Sometimes the cat can stir up the dirt to accelerate the return of native plants. If it isn't used carefully, it can bring weeds or accelerate erosion as you suggest. It all depends upon circumstances and methods.
I think the emphasis would be on the stabilizing effects of ground cover root networks. Obviously soil aeration and chemistry would be important to healty root growth.
The biggest risk of sedimentation in streams is if we DONT thin the stands. If the forest burns too hot in a cataclysmic crown fire, the trees WILL die to a greater degree than if it had been clearcut. It will be no mosaic burn; the disturbed area will be huge. There will be no surface plants to slow the water. There will be no duff to filter the soils. When it rains, the suspended solids will act like abrasive slurry to cut the soil and destabilize slopes. There will be 0% canopy for nearby streams, but then they will likely be so full of mud it wont matter to the fish.
I absolutely agree. The 50+ year policy of "no burn" has resulted in way too much flammable deadfall on the forest floor. That's why the Yellowstone fires were as drastic as they were. "Normal" forest fires don't get so hot that they result in much crown burns -- they burn the litter on the forest floor. That's why I think there should be a major effort to clear deadfall and then initiate a controlled burn plan over much of the nation's forest. But certainly not until the drought abates (though deadfall clearing should get underway as soon as possible).
Your picture is illustrative. One thing that is perhaps left out of your discussion is the role of browsers in trimming shoots so that the natural density of clusters from stumps is reduced. By thinning clusters, you perform the same function.
Thanks for an illustrative reply.
Deer don't eat redwood willingly, there is too much other vegetation that they greatly prefer (such as ceanothus).