ESTIMATING CLIMATE FROM THE LATE CRETACEOUS FLORA OF THE TWO MEDICINE FORMATION, MONTANA: COMPARISON OF UNIVARIATE METHODS AND DIGITAL LEAF PHYSIOGNOMY

Plants are directly affected by climate. Leaf architectural features that are useful for climatic reconstruction include leaf size, type of margin, and venation. Starting with the research of Bailey and Sinnott in 1915, scientists have worked to relate these features to climate. Several key analytical methods have been developed for analyzing extant and fossilized leaves, including Leaf Margin Analysis (LMA) for inferring Mean Annual Temperature (MAT), and Leaf Area Analysis (LAA) for inferring Mean Annual Precipitation (MAP). These two methods have been augmented by the multivariate methods of Climate Leaf Analysis Multivariate Program (CLAMP) and Digital Leaf Physiognomy (DiLP). DiLP utilizes digital editing software to measure different leaf features, and can be used to predict both temperature and precipitation.

This study uses LMA, LAA, and DiLP to reassess the paleoclimate inferred from the fossil flora of the Upper Cretaceous Two Medicine Formation of Montana. The specimens used in this study are twenty-nine morphotypes of fossil dicotyledon (dicot) leaves collected from the lower part of the formation near Cut Bank, Montana by David Crabtree. Crabtree estimated MAT of 7–10 °C with no extended cold season, based on LMA and the presence of fossil palms.

In this analysis, DiLP estimates a MAT of 17°C, whereas LMA estimates a lower temperature of 10°C. DiLP and LAA estimate MAP of 108 and 91 cm, respectively, values characteristic of a subhumid climate. The temperature offset between DiLP and LMA for the Two Medicine flora is comparable to temperature offsets found for Maastrichtian and Paleocene floras from the northern Western Interior. The DiLP estimate of MAT is more in line with published temperature estimates for the Campanian of North America, which may indicate that it provides the more accurate estimate of temperature for the Two Medicine flora.