SIZING UP THE LEAVES OF AN EOCENE PATAGONIAN PALEORAINFOREST AND ITS AUSTRALIAN ANALOGS
Fossil leaf sizes were measured directly or estimated from fragmented specimens. Intact leaves were used to test three methods of leaf area estimation: the Cain & Castro formula, the Raunkiaer-Webb size categories, and a new technique applicable to more fragmented leaves that uses global negative scaling relationships between 2° vein density and leaf area (Sack et al. 2012). Areas determined with the Cain & Castro formula fit best with actual values, and those from the Raunkiaer-Webb categories and 2° vein scaling gave comparable results of lower accuracy. All methods were applied to LH specimens, producing datasets of four confidence levels for a robust characterization of assemblage-wide and taxon-specific leaf size. Modern analog forests will be selected as those with similar leaf sizes and forest structure to LH, as inferred from the life forms of well-described fossil taxa and photosynthetic productivity suggested by leaf vein density. The most similar potential analog forests will be statistically compared to discover common ecological features, such as canopy closure and soil drainage, that may be representative of the LH paleorainforest. Reconstructing LH will refine our understanding of a mid-latitude, extinct rainforest biome that was present during the highest sustained global temperatures of the Cenozoic. This new method of using leaf size for ecosystem reconstruction can be used on many other paleofloras with significant numbers of close living relatives.