Soil- and plant-water dynamics in a C3/C4 grassland exposed to a subambient to superambient CO2 gradient

Authors

  • H. WAYNE POLLEY,

    1. Grassland, Soil and Water Research Laboratory, U.S. Department of Agriculture, Agricultural Research Service, Temple, Texas 76502, USA
    Search for more papers by this author
  • HYRUM B. JOHNSON,

    1. Grassland, Soil and Water Research Laboratory, U.S. Department of Agriculture, Agricultural Research Service, Temple, Texas 76502, USA
    Search for more papers by this author
  • JUSTIN D. DERNER

    1. Grassland, Soil and Water Research Laboratory, U.S. Department of Agriculture, Agricultural Research Service, Temple, Texas 76502, USA
    Search for more papers by this author

: Wayne Polley, fax +254 770 6561, e-mail: polley@brc.tamus.edu.

Abstract

Plants may be more sensitive to carbon dioxide (CO2) enrichment at subambient concentrations than at superambient concentrations, but field tests are lacking. We measured soil-water content and determined xylem pressure potentials and δ13C values of leaves of abundant species in a C3/C4 grassland exposed during 1997–1999 to a continuous gradient in atmospheric CO2 spanning subambient through superambient concentrations (200–560 µmol mol2−1). We predicted that CO2 enrichment would lessen soil-water depletion and increase xylem potentials more over subambient concentrations than over superambient concentrations. Because water-use efficiency of C3 species (net assimilation/leaf conductance; A/g) typically increases as soils dry, we hypothesized that improvements in plant-water relations at higher CO2 would lessen positive effects of CO2 enrichment on A/g. Depletion of soil water to 1.35 m depth was greater at low CO2 concentrations than at higher CO2 concentrations during a mid-season drought in 1998 and during late-season droughts in 1997 and 1999. During droughts each year, mid-day xylem potentials of the dominant C4 perennial grass (Bothriochloa ischaemum (L.) Keng) and the dominant C3 perennial forb (Solanum dimidiatum Raf.) became less negative as CO2 increased from subambient to superambient concentrations. Leaf A/g—derived from leaf δ13C values—was insensitive to feedbacks from CO2 effects on soil water and plant water. Among most C3 species sampled—including annual grasses, perennial grasses and perennial forbs—A/g increased linearly with CO2 across subambient concentrations. Leaf and air δ13C values were too unstable at superambient CO2 concentrations to reliably determine A/g. Significant changes in soil- and plant-water relations over subambient to superambient concentrations and in leaf A/g over subambient concentrations generally were not greater over low CO2 than over higher CO2. The continuous response of these variables to CO2 suggests that atmospheric change has already improved water relations of grassland species and that periodically water-limited grasslands will remain sensitive to CO2 enrichment.

Ancillary