You are here:
HYDRAULIC REDISTRIBUTION OF SOIL WATER: ECOSYSTEM IMPLICATIONS FOR PACIFIC NORTHWEST FORESTS
Citation:
Brooks, J R., F. C. Meinzer, J. C. Domec, J. M. Warren, AND R. Coulombe. HYDRAULIC REDISTRIBUTION OF SOIL WATER: ECOSYSTEM IMPLICATIONS FOR PACIFIC NORTHWEST FORESTS. Presented at Ecosystems Analysis Series (University of Washington), Seattle, WA, June 5, 2003.
Description:
The physical process of hydraulic redistribution (HR) is driven by competing soil, tree and atmospheric water potential gradients, and may delay severe water stress for roots and other biota associated with the upper soil profile. We monitored soil moisture characteristics across three forest ecosystems in the Pacific Northwest to investigate the spatial and temporal variability of HR of water by roots. Soil water potential and volumetric water content were measured down to depths of 1 or 2 m, respectively, in an old growth (300+ years) and young (20 years) ponderosa pine stand, and in old growth (500+ years) and young (25 years) Douglas-fir stands. Significant diel fluctuations in soil water content and water potential indicated that HR occurred in all four ecosystems. Spatial variability of HR to the upper soil layers (20-60 cm) was large within each site. Initiation and continuance of HR were dependent upon reaching a threshold soil water potential of about -0.2 MPa (-0.05 to -0.6 MPa). Hydraulic redistribution occurred earlier in the season at the drier pine sites, and was induced at higher water potential thresholds than in the Douglas-fir stands.
In addition, we manipulated the system by irrigation, soil trenching, and tree removal to alter competing water sources and sinks. In the irrigation experiment, we applied 2100 l of 9000 l deuterated water to a 1 m2 plot over three weeks, and sampled soil and the surrounding vegetation for 36 days. Deuterated water was immediately taken up by the surrounding dominant 25 yr-old Douglas-fir trees and within 7 days was detected in surface soils in front of those trees, but at least 1 m from the watering site. By the end of the sampling period (36 days) deuterated water was detected in soils behind the target trees, and in Oregon grape, blueberries, and small understory hemlocks as much as 4 m away from the watering source. The amount of HR in the upper soil layers at the watering site was twice that of the control, and the amount of water utilized from the upper soil was also increased. Trenching was hypothesized to decrease HR by separating a section of soil from the trees that had access to ground water; but because a large root in the plot was severed from a nearby dominant Douglas-fir tree, HR and soil moisture actually increased. We assumed that once the large sink of the tree was removed from the root, the soil sink received more water from the root. To confirm this, we measured sapflow on several roots on another tree, and then cut down the tree. In some roots we found a dramatic reversal of flow from up the tree to flow out to the surrounding soil. These manipulations highlighted that the magnitude of HR is governed by a competition between shallow roots in dry soil and the aboveground portion of the tree for water taken up by deep roots. Hydraulically redistributed water can move several meters from the original source where it can be utilized by other vegetation.