Numerous paired watershed studies have addressed the role of forests in regulating water yield and have demonstrated the tight interaction between vegetation and streamflow production. New challenges, such as the effects of non-random species loss on water yield, are emerging, and are already influencing forested ecosystems in the eastern US. In particular, eastern hemlock (Tsuga canadensis (L.) Carr.) is one of the principal riparian and cove canopy species in the southern Appalachian Mountains. Throughout its range, eastern hemlock is facing potential widespread mortality from hemlock woolly adelgid (HWA; Adelges tsugaeAnnand). We used a paired watershed approach to estimate the impact that the loss of hemlock will have on streamflow. Monthly streamflow records from two closely–located watersheds similar in size, aspect, elevation, but differing substantially in pre-management cover conditions were used to estimate the impact of hemlock before (2000–2004) and after (2005–2010) HWA infestation on the hydrologic budget. We hypothesized that hemlock loss would result in increased annual streamflow (ca. 10%), and that the timing of the streamflow increase would not be evenly distributed, with winter and spring experiencing greater increases compared to summer and fall.
Results/Conclusions
Prior to HWA infestation, monthly streamflow relationships between reference and HWA infested watersheds were highly significant (R2 > 0.93, P < 0.05), except for Sep and Oct, in which either the replication was low due to weir repairs during 2003 (R2 = 0.96, P < 0.12), or the spread of the data was insufficient (R2 = 0.86, P = 0.07), respectively. After infestation, relationships between reference and infested watersheds appeared to be better described by a different linear model. For all months except Jun and Nov, as streamflow increased in the reference watershed, a greater increase in the HWA infested watershed resulted. After HWA infestation, we predicted what monthly streamflow for the disturbed watershed would have been under the given climate conditions had HWA infestation not occurred. We observe both higher and lower streamflow than expected after HWA infestation. In general these departures were all within the bounds of prediction error; however, some months and years had significant departures. After summing all the significant departures in streamflow from expected, we found higher streamflow than expected by approximately 3–5% of total annual streamflow and 3–7% of annual evapotranspiration during years 2005, 2009 and 2010.