Carbon dioxide as a selective agent for stomatal density

Chantal D. Reid1, Robert B. Jackson1, and Joy K. Ward2

1Department of Biology, Duke University, Durham, NC, USA 27708
2Department of Biology, University of Utah, Salt Lake City, UT, USA 84112


In long-term exposure to elevated CO2, adaptive responses may include a change in stomatal density as well as a change in leaf conductance. No study to our knowledge includes a test, under controlled environmental conditions, of such an evolutionary selection although several reports indicate that stomatal density may change with atmospheric CO2 (e.g. herbarium specimens). To elucidate if stomatal density is affected by selection at different [CO2], we grew Arabidopsis thaliana at pre-industrial, ambient, and elevated atmospheric CO2; 800 µE m-2 s-1 PAR. In a first experiment, we used seeds from a previous field experiment where Arabidopsis plants having low and high stomatal density were identified relative to the average population. The plants were grown at 22° C day: 18° C night as in the field experiment. In a second experiment, we used seeds from a previous selection experiment where Arabidopsis plants were selected for high seed yield at different [CO2] over 5 generations. In both experiments, the different populations generally increased their stomatal density from low to ambient CO2 rather than decrease as expected. Yet, stomatal indices declined with increasing CO2 as the epidermal cells increased. The decline in leaf area with increasing CO2 observed in the first experiment may explain this unexpected trend. A lack of change in leaf conductance for plants selected at 200 µmol CO2 mol-1 in all growth CO2 in the second experiment suggests that stomatal function was altered by inadvertent selection by CO2.

Leaf surface impressions at 400 x magnification

Plants grown at 260 µmol CO2 mol-1

High stomatal density population

Low stomatal density population



  • Stomatal conductance depends on the density of stomata on the leaf and on their aperture opening, which controls diffusion through individual stomata.
  • Stomatal conductance is decreased by elevated atmospheric [CO2] (e.g., Morison, 1985; Mansfield et al., 1990; Drake et al., 1997).
  • Stomatal density is often altered with elevated CO2 although opposite trends are reported. See Table 1.
  • Decrease in stomatal density under elevated CO2 can be heritable (Case et al., 1998).


Elevated CO2 will be a selective agent to reduce stomatal density because it reduces leaf conductance. Likewise, low CO2 will act as a selective pressure to increase stomatal density.


To determine if different [CO2] can alter the stomatal density and index as well as the stomatal function of Arabidopsis after several generations of growth.


  • Plant material: Arabidopsis thaliana
  • Experiment 1: Stomatal selection
    • Seeds from populations of low, control, and high stomatal density (208, 407, 545 stomata mm-2, respectively) were used (Reid and Fiscus, unpubl.) and grown in C chambers at the Duke Phytotron.
    • Plants were grown in Metromix-200 in flats at 100 plants per flat.
    • Watering and nutrient applications were by sub-irrigation.
    • Temperature: 22°C/18°C for 13/11 h light/dark
  • Experiment 2: Inadvertent selection
    • Seeds from populations grown at 200 or 700 µmol CO2 mol-1 for 5 generations and selected for high seed mass (Sel) or without selection (Ctr) were used after growth in a common garden. They were grown in C chamber at the Duke Phytotron.
    • Grown in 750-ml pots in gravel:Turface:Vermiculite (1:1:1) at 1 plant per pot
    • Watering and nutrient were by misting
    • Temperature: 25°C/18°C for 14/10 h light/dark
    • All watered twice a day and fertilized once a day with 1/2 Hoagland’s solution
    • Stomatal impressions:
    • Casts of the leaves were made using dental impression material (Kerr Extrude-Medium).
    • Leaf impressions were made from clear nail polish peels of the cast.
    • Number of stomata and epidermal cells were recorded under a light microscope hooked up to a computer imaging system.


Stomatal selection

Number of stomata increased from low to ambient CO2 and stomatal index decreased. The high density population was most responsive.

  • The number of stomata increased significantly from low to ambient CO2 only (CO2 effect at P<0.001), in contrast to our hypothesis.
  • The number of epidermal cells increased significantly with CO2 (P<0.001) except for the low stomatal density population (CO2 x density at P<0.0001).
  • Because of the large increase in epidermal cells, the stomatal index was decreased from low to ambient CO2 in all and from low to elevated CO2 for the high stomatal density population (CO2 effect at P<0.001, CO2 x density at P<0.0001)

Inadvertent selection

[CO2] does not cause a directional selection for stomatal density or function as both populations selected at 200 and 700 ppm showed similar response to growth CO2.

  • Although small, the number of stomata increased with CO2 (P<0.02) mostly because of the unselected populations.
  • The number of epidermal cells increased significantly with CO2 (P<0.002) mostly in the unselected populations.
  • The stomatal index was decreased as CO2 increased in the unselected populations (P<0.007)


Stomatal selection

The decrease in leaf area with CO2 was correlated with the increase in number of stomata, suggesting that change in leaf size was responsible for the change in stomata. Increased competition as CO2 increased may have reduced leaf size.


Inadvertent selection

Plants selected at 200 µmol CO2 mol-1 have lost their functional response to CO2 and the response in unrelated to stomatal density or index.

Leaf conductance was measured with LiCor 6400. Data are shown for plants grown and measured at growth CO2



  • The change in stomatal density with CO2 may be via an effect on total leaf area, although stomatal index decreases as CO2 increases.
  • Low CO2 may act as an inadvertent selective agent for decreased stomatal function



Research was funded by a DOE grant to RB Jackson. Thanks to Jessica Odom and Catarina Moura for counting stomates, Will Cook for gas exchange.


This poster was presented at the 2000 Ecological Society of America meeting.

Current Projects | Jackson Lab home

Last modified 9 October 2000