Leaf stomata facilitate the exchange of water and CO2 during photosynthetic gas exchange. The shape, size, and density of leaf pores have not been constant over geologic time, and each morphological trait has potentially been impacted by changing environmental and climatic conditions, especially by changes in the concentration of atmospheric carbon dioxide. As such, stomatal parameters have been used in simple regressions to reconstruct ancient carbon dioxide, as well as incorporated into more complex gas-exchange models that also leverage plant carbon isotope ecology. Most of these proxy relationships are measured on chemically cleared leaves, although newer techniques such as creating stomatal impressions are being increasingly employed. Additionally, many of the proxy relationships use angiosperms with broad leaves, which have been increasingly abundant in the last 130 million years but are absent from the fossil record before this. We focus on the methodology to define stomatal parameters for paleo-CO2 studies using two separate methodologies (one corrosive, one non-destructive) to prepare leaves on both scale- and broad-leaves collected from herbaria with known global atmospheric CO2 levels. We find that the corrosive and non-corrosive methodologies give similar values for stomatal density, but that measurements of stomatal sizes, particularly guard cell width (GCW), for the two methodologies are not comparable. Using those measurements to reconstruct CO2 via the gas exchange model, we found that reconstructed CO2 based on stomatal impressions (due to inaccurate measurements in GCW) far exceeded measured CO2 for modern plants. This bias was observed in both coniferous (scale-shaped) and angiosperm (broad) leaves. Thus, we advise that applications of gas exchange models use cleared leaves rather than impressions.
Keywords: atmosphere; carbon dioxide; paleobarometer; paleoclimate; photosynthesis; stomata.