The contribution of deep roots to the water cycle: seasonal patterns of water uptake and hydraulic lift from below 7 m on the Edwards Plateau, Texas

Andrew J. McElrone1, William T. Pockman2, and Robert B. Jackson1

1Department of Biology, Duke University, Durham, NC 27708 2Department of Biology, University of New Mexico, NM 87131

Abstract

The presence of plant roots in deep and shallow soil is potentially important for site water balance through water uptake and redistribution between compartments. However, direct measurements of water movement through deep roots are not usually possible. We used cave systems on the Edwards Plateau in central Texas to gain access to deep roots of Juniperus ashei and Quercus fusiformis between 7 and 20 m below the soil surface. Long-term sap flux measurements of J. ashei showed that a single large root measured 7 m below the surface contributed 20 – 50% of daily transpiration, depending on the water content of surface soils. During periods without rain, upward flow through deep roots was continuous during both day and night with nocturnal hydraulic lift contributing up to 20% of daily water movement from depth. Seasonal measurements of roots of Q. fusiformis drawing upon an underground stream 20 m below the surface showed a similar pattern. In both systems, sap flux through deep roots fell to zero during the night when the water content of surface soils was high following precipitation. As surface soil water content decreased, minimum flow rate through deep roots increased steadily before dropping back to zero when surface soil water content rose following precipitation. Our results show that these large woody trees play an important role in cycling water from depth in these karst systems.

Figure 1: Left to Right. Q. fusiformis tree utilizing perennial underground stream; Cotterell Cave entrance, Austin, TX; cave datalogger (DL) used for belowground measurements; sap flux and psi sensors on deep tap roots.

Background

Importance of Deep Roots

  • Deep roots are found in most biomes, especially water-limited ones
  • The abundance of deep-rooted plants is increasing in rangelands globally with woody plant encroachment
  • Deep roots contribute substantially to whole tree and ecosystem water use via uptake and hydraulic redistribution
  • Caves can provide access to intact, functioning deep roots for in situ measurements

Objectives

  1. Understand the contribution of deep roots to whole tree water use
  2. Assess seasonal changes in deep root water uptake and hydraulic lift

Methods

Sites:
Cotterell Cave in Austin, TX. A limestone cave with large Ashe juniper roots (J. ashei) passing through the cave at ~7 m depth.
Powell’s Cave in Menard, TX. A large limestone cave with Woolly buckthorn (Bumelia lanuginosa), Live Oak (Quercus fusiformis), and Sugar hackberry (Celtis laevigata) roots penetrating to ~20 m depth to access a perennial underground stream.

Root ID: Established root-shoot connections of individual trees with DNA fingerprinting techniques (AFLP and ISSR).

Sap Flux: Heat balance, Heat ratio method and Granier sensors.


Figure 10: Left: J. ashei deep root in Cotterell Cave; Right: another day at the grind for W.T. Pockman in Powell's Cave.

Conclusions

Deep root water use varies seasonally with changes in environmental parameters.

Shoot and deep root responses are tightly linked suggesting little water storage. Additionally, we have found no evidence of nighttime transpiration. Water is likely being hydraulically redistributed to dry shallow soil or is replenishing daily transpired water.

Nighttime water flow in deep roots occurs in both species. Availability in aboveground, shallow soil moisture alters deep root water use patterns. Specifically, precipitation events wet the shallow soil and change water potential gradients, thus temporarily eliminating the driving gradient for hydraulic lift until the shallow soil water is depleted. Alternatively, decreased shoot water use on days with precipitation decreases the need for nighttime replenishment from depth.

Cave systems like the one utilized for this study provide an opportunity to understand how shallow and deep portions of the root system differ in structure, function and response to changing environmental conditions.

Acknowledgements

Thanks to:

James Powell for access to Powell’s Cave and Mark Sanders for access to Cotterell Cave; Shawn Brumbaugh, Brian McElrone, Marcy Litvak, Vic Engel, Tracey Crocker and Cynthia Willson for field assistance; Bill Cable and Jim Kjelgaard for advice on heat balance sensors; Randy Linder for DNA fingerprinting deep roots. This work was funded by: USDA-NRI, NSF and Mellon Foundation grants.

 

This poster was presented at the 2003 AGU (American Geophysical Union) Fall meeting.

Current Projects | Jackson Lab home

Last modified 9 April 2004 cwcook@duke.edu