Historical Records of Stomatal Indicies in Quercus from the Southeastern U.S
Document Type Honors Thesis (Open Access)
Modeling atmospheric carbon dioxide concentration (pCO ) is critical to understanding the Carbon Cycle over geologic time. Recently, biological proxies have become commonly used including stomatal frequencies on leaves. An inverse relationship exists between epidermal stomata and the pCO under which the leaf grew and expanded; this provides a record of atmospheric gases as accurate as ice core data, another proxy, but at a higher temporal resolution. The two most commonly used leaf parameters are Stomatal Density (SD - # of stomata/unit area) and Stomatal Index (SI -the ratio between # stomata and total epidermal cells/unit area). However, stomatal frequencies have some limitations. A CO “ceiling” exists where plants stop responding to gas concentration in a linear fashion above a certain threshold. Additionally, transfer functions calibrated from extant plants do not always correspond to fossil equivalents. Another recent consideration is that taxonomically related plants may not exhibit similar growth responses under the same pCO conditions. To test this, 12 species of Oak (Quercus) were evaluated with respect to SD and SI over a 104 - year interval based on trees grown under a humid subtropical climate in Lee County, AL. Herbarium specimens dating from 1894 to the 1980s were used to supplement the historical record; a collection made during June 2007 extended the record to the present. Herbarium samples were taken from collections made in the same year, if available. Materials were cleared in chromic acid, mounted on slides, after where stomata and epidermal cells were counted using a Zeiss Axioskop and AxioVision software. Results appear to demonstrate that Oak species respond independently to pCO . Both sections of Quercus, white and red, have statistically different responses to pCO when results were compared. Some species, such as Q. laurifolia and Q. nigra, have similar responses over time, but all responses tested are statistically unique. Therefore, the present study suggests that approaches correlating data from extant taxa with data from extinct taxa are debatable and should be considered for their qualitative observations rather than quantitative measures.