You are here

Harvard Forest >

Harvard Forest Symposium Abstract 2005

  • Title: Water use by eastern hemlock (Tsuga canadensis) forest and deciduous forests in central Massachusetts: Hydrologic implications
  • Primary Author: Julian Hadley (Harvard Forest)
  • Additional Authors: Paul Kuzeja (Harvard Forest)
  • Abstract:

    Water use by forests strongly influences the amount of precipitation that enters streamflow and groundwater. To better understand differences in water use between forest types, in the spring of 2004 we began to measure evapotranspiration (ET), or water vapor transfer from land to atmosphere, in a 100 to 200 year old eastern hemlock (Tsuga canadensis) forest in the Prospect Hill section of the Harvard Forest. We measured ET from the hemlock forest and from a younger deciduous forest about 600 m to the northwest by the eddy covariance technique, which enables us to estimate both water and carbon exchange of each forest type. Both of these forests have an average leaf area index (LAI) between 4 and 4.5 during the summer. In late June and early July, the hemlock forest lost about 2 mm of water per day to the atmosphere, about half the amount of water lost to the atmosphere by deciduous forest. However, during the summer ET from deciduous forest declined more rapidly than ET from hemlock forest, so that by September hemlock forest was about two-thirds of deciduous forest ET. Deciduous forest ET fell to near zero after leaf fall in mid-October (Figure 1). From mid-June to mid-September, total ET from deciduous forest was about 250 mm, nearly equal to total precipitation measured at the Harvard Forest weather station. In contrast ET from hemlock forest was about 100 mm less (Figure 2). Nearly 250 mm of additional precipitation in late September and October was far greater than ET by either forest type.


    Higher ET from deciduous compared to hemlock forest was associated with lower surface soil moisture in the deciduous forest during the summer (Figure 3). Light, or more specifically photosynthetically active radiation (PAR), exerted the strongest environmental influence on ET in both hemlock and deciduous forests. The slope of the relationship between ET and PAR in early summer was twice as high for deciduous forest as for hemlock forest (Figure 4). PAR and water vapor pressure deficit of the air (VPD) were used to estimate ET when wind direction did not allow ET measurements, and during occasional instrument failures.


    These results imply that significant changes in streamflow and soil water content during summer will result from the death of hemlock forests caused by the hemlock woolly adelgid. This insect has already caused large-scale hemlock mortality in Connecticut (Orwig and Kizlinski 2002) and it is present in small numbers at Harvard Forest. Thus far, the primary species replacing hemlocks killed by the woolly adelgid is black birch (Betula lenta). Measurements in 2003 showed that leaf conductance to water vapor in black birch was about twice as high as in eastern hemlock. Leaf conductance in black birch was similar to red oak (Quercus rubra), the dominant species in the deciduous forest for which ET was measured in this study. Therefore, when black birch trees are large enough to have a total leaf area similar to the hemlocks that they replace, streamflow and soil water are likely to be substantially lower than in the previous hemlock-dominated forest.





    Reference:


    Orwig, D.A. and M.L. Kizlinski. 2002. Vegetation response following hemlock woolly adelgid infestation, hemlock decline, and hemlock salvage logging. pp. 106-117 In Proceedings: Hemlock Woolly Adelgid in the Eastern United States Symposium. (eds. R.C. Reardon, B.P. Onken, and J. Lashomb). New Jersey Agricultural Experiment Station Publication, New Brunswick, NJ.


  • Research Category: