Seasonal and successional streamflow response to forest cutting 2 and regrowth in the northwest and eastern United States

6 [1] This study examined daily streamflow response over up to four decades in northwest 7 conifer forest and eastern deciduous forest sites in the United States. We used novel 8 methods to analyze daily observations of climate and streamflow spanning more than 9 900 basin years of record at 14 treated/control basin pairs where forest removal and 10 regrowth experiments were underway in the period 1930–2002. In the 1 to 5-year period 11 after forest removal, maximum daily increases ranged from 2 to 3 mm at deciduous forest 12 sites, to 6 to 8 mm at conifer forest sites. Significant spring surpluses persisted for up 13 to 35 years in conifer forest basins, but winter and spring streamflow deficits appeared 14 after 10 to 15 years of forest regrowth in eastern deciduous forest basins. In all 5-yr 15 posttreatment periods, absolute changes in daily streamflow were significantly more likely 16 during moist, warm seasons, or during snowmelt seasons, but relative changes were more 17 likely during warm seasons irrespective of moisture status. Both relative and absolute 18 streamflow changes in the 1 to 5 and 15 to 25-year periods after forest removal were 19 significantly positively related to the age of the forest at the time it was cut. Eastern 20 deciduous forests had been disturbed by logging or hurricane 12 to 56 years prior to forest 21 removal, while Pacific Northwest conifer forests had been not experienced logging or 22 wildfire for 90 to 450 years. Paired basin experiments provide a continuous, and 23 continuously changing, record of vegetation structure, composition, and climate, and their 24 effects on streamflow.