Publication Date

April 2017


Dana Royer


Earth & Environmental Sciences


English (United States)


Many geological and paleontological studies depend on reliably estimating ancient climates, and atmospheric carbon dioxide concentration is a key parameter in our climate system. Plant stomata can be used as CO2 proxies, but the existing proxy methods have many shortcomings. Franks et al. (2014) recently developed a new stomatal CO2 proxy based on fundamental principles of plant physiology. This method requires measuring only stomatal size and density and the d13C of the leaf tissue. This means it can easily be used with fossils of any stomatal-bearing plant species because on most plant fossils stomata can be observed and d13C can be measured. However, the Franks model has not been extensively tested for accuracy and sensitivity. This study provides a rigorous test of the model by measuring only the required parameters on just one individual of thirty-one different species. The species tested represent a wide range of taxonomic orders, plant types and growth forms. I find that among arborescent species the model produces relatively accurate estimates, with the average estimate only 50 ppm higher than the target CO2 concentration. I do not detect a phylogenetic bias, including among angiosperms versus gymnosperms or between taxonomic orders. Herbaceous species produce higher estimates on average, but this can be accounted for by the fact that the CO2 environment which these species experience on the forest floor is quite different from that of arborescent species. I show that using a two-endmember mixing model to determine a more appropriate air d13C value improves CO2 estimates for herbaceous species significantly. This study also tests the sensitivity of most of the key parameters in the Franks model. I find that leaf d13C, stomatal pore length and stomatal density should be carefully measured with replication from multiple individuals of the same species if possible because the model is quite sensitive to these parameters, while it is less sensitive to many of the inputs that can be substituted with values recommended by Franks et al. (2014), such as the reference ratio between intercellular and atmospheric CO2, air d13C and reference photosynthetic rate. The Franks model is a great improvement on previous stomatal CO2 proxies, and should be able to provide sufficiently accurate estimates of paleo-climates if adequate replication among fossil individuals and species is achieved.



© Copyright is owned by author of this document