Tree Physiology 31, 649–658 doi:10.1093/treephys/tpr052
Research paper
The fate of hydraulically redistributed water in a semi-arid zone eucalyptus species Kim Brooksbank1,2,4, Erik J. Veneklaas1,2, Donald A. White1,2,3 and Jennifer L. Carter2,3 1School
of Plant Biology, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia; 2CRC for Future Farm Industries, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia; 3CSIRO Sustainable Ecosystems, CSIRO Centre for Environment and Life Sciences, Private Bag 5, Wembley, WA 6913, Australia; 4Corresponding author (Email:
[email protected]) Received February 16, 2011; accepted May 11, 2011; published online July 8, 2011; handling Editor David Whitehead
Keywords: deuterium, efflux, groundwater use, hydraulic redistribution, salinity.
Introduction Competition for water and nutrients between plants in arid environments is intense (Gibbens and Lenz 2001), resulting in the need to maximize access to soil water while minimizing the cost to the plant in terms of root biomass allocation (Gries et al. 2003). The root system architecture has a significant influence on patterns of soil water utilization by trees and ultimately canopy transpiration (Ong et al. 2002). One way in which plants have been shown to improve access to limited soil water resources is through hydraulic redistribution (HR) (Burgess et al. 1998), which describes transport of water via roots along water potential gradients from wetter to drier parts of the soil profile (Richards and Caldwell 1987). This usually occurs nocturnally when reduced transpiration results in the root water potential (ψ x) in dry soil layers rising above the water potential of the surrounding soil (ψs) (Hultine et al. 2003). The root system may traverse multiple soil layers of differing moisture content. The ability of a tree to relocate water
from one soil horizon to another or even one location to another within a particular soil horizon has been reported in a number of species around the world (Burgess et al. 1998, Meinzer et al. 2001, Scholz et al. 2010). A 1998 review reported that HR had been confirmed in 27 species of grasses, herbs, shrubs and trees (Caldwell et al. 1998), but noted that the phenomenon was likely to occur wherever plants experienced gradients in soil moisture between different parts of the root system. Recent work has provided evidence of HR occurring in Eucalyptus kochii subsp. borealis (Brooksbank et al. 2011). Confirmation of this process in arid zone eucalypts will have implications for agroforestry design principles in dryland agricultural systems where water availability is an issue. Species with a root architecture that facilitates HR may increase total stand water use through efflux of water from surface roots (Caldwell 1990), and efflux of water from surface roots following HR has been documented (Burgess et al. 2000a, 2000b, Domec et al. 2004, Warren et al. 2007). It appears that significant amounts of water can move in this way.
© The Author 2011. Published by Oxford University Press. All rights reserved. For Permissions, please email:
[email protected]
Downloaded from treephys.oxfordjournals.org by guest on July 27, 2011
Although hydraulic redistribution has been observed for a range of tree species, including Eucalyptus kochii subsp. borealis (C. Gardner) D. Nicolle, there is limited direct evidence that water taken up by deep roots in moist soil is in fact exuded by shallow roots in dry soil. This paper reports an experiment designed to test this hypothesis. Water enriched with deuterium was added to the groundwater via a slotted tube at 4.5 m depth below 5-year-old E. kochii subsp. borealis trees. Nocturnal sap flow increased markedly immediately after deep irrigation, indicating that the trees were using water from this depth. Two weeks later, samples of surface soil and xylem water were found to contain levels of deuterium up to 30% higher than soils and xylem water from a control plot upslope of the main treatment plot. This is strong evidence that trees used groundwater and that efflux of important amounts of hydraulically redistributed water occurred via the roots of E. kochii subsp. borealis.
650 Brooksbank et al. two-row belts oriented north–south to reduce erosion of the sandy soil by the prevailing winds. E. kochii is a mallee form (multi-stemmed) eucalypt native to this area. Within the belts the rows were planted 2 m apart, and within rows trees were also 2 m apart. The belts were established ~110 m apart. The site has a dry Mediterranean-type climate with a long-term (100-year) mean annual rainfall of 380 mm, 79% of which falls between May and October. Long-term (100-year) average Priestley–Taylor evapo transpiration peaks at ~350 mm a month in the dry season and reduces to ~70 mm a month in the wet season. Long-term average total annual evapo transpiration is 2285 mm. The soils are 4–6 m of pale yellow sands consisting of 80% quartz with the remainder goethite and kaolinite. This overlies a relatively impermeable granitoid saprolite or silcrete (George 1992). Relatively fresh groundwater (electrical conductivity of
