Stomata as CO2 Sensors
In 1987, F. Ian Woodward published a seminal paper showing that stomatal density on herbarium specimens of temperate trees had declined measurably since the 18th century — tracking the industrial rise in atmospheric CO2. This discovery launched a new paleoclimate proxy: if modern plants adjust their stomatal numbers in response to CO2, then fossil leaves should record ancient CO2 levels in their epidermal anatomy.
The Stomatal Index
Raw stomatal density (stomata per mm2) varies with cell size, which changes with light, water, and leaf developmental position. The stomatal index — the fraction of epidermal cells that are stomatal — normalizes for cell size differences and provides a more robust CO2 proxy. Under a microscope, counting a few hundred cells in a standardized area of cuticle gives an SI value accurate to about 0.5 percentage points.
Calibration Through Geological Time
Ginkgo biloba is the gold standard for deep-time CO2 reconstruction because the genus has existed for over 200 million years with remarkably conserved leaf morphology. By comparing fossil Ginkgo SI values to the modern SI-CO2 relationship, researchers have reconstructed CO2 through the Mesozoic and Cenozoic. The results agree broadly with independent proxies (boron isotopes, alkenones, paleosol carbonates), supporting CO2 levels of 1000-2000 ppm during the Cretaceous greenhouse.
Limitations and Modern Extensions
The stomatal proxy saturates at high CO2 — above about 800-1000 ppm, plants cannot reduce stomata much further, creating a ceiling on sensitivity. Cuticle preservation quality matters: only well-preserved adaxial or abaxial surfaces with clear cell outlines yield reliable counts. Machine learning approaches are now being developed to automate stomatal counting from microscope images, dramatically increasing throughput and reproducibility.