Sometimes when a forested watershed burns (partially) down, or blows over, water yield from that watershed increases. That's potentially good - more water - and potentially bad - floods. Other times yield decreases, or occurs earlier/later.
A publication of mine and Ben Livneh's (University of Colorado) is now appearing in Environmental Research Letters, where we explore how the context of a landscape (e.g., it's elevation, climate, topography, and composition) influence its water yield response to disturbances. Larger disturbances almost always lead to more water yield from a forest, though there is some context on how major the increase is - some higher precipitation watersheds didn't see quite the proportional increase in water that other areas did, for example. These results certainly lead to more questions - what about finer scale variation (e.g., just dry watersheds in the west?). But they also enable a first stab at deciding what watersheds and water systems are sensitive to future disturbances, and which are likely pretty resistant to change, regardless of disturbance. Good to know if you care about the water coming down the mountain in your neighborhood.
Buma B., Livneh B. Key landscape and biotic indicators of watersheds sensitivity to forest disturbance. Environmental Research Letters. In press.