ESTIMATING PALEOALTITUDE FROM FOSSIL PLANTS: SOURCES OF ERROR

Estimates of paleoaltitudes for Tertiary leaf assemblages from continental interiors have generally been based on comparisons between interior assemblages and isochronous assemblages from near-coastal positions at the same latitudes. For a given assemblage, this method assigns mean annual temperatures (MAT) to both interior and coastal assemblages and assumes a general terrestrial (environmental) lapse rate (TLR), i.e., a general decrease in MAT with altitude at a given latitude. Many estimates of paleoaltitudes further rely on assigning MAT based on the present-day tolerances of the nearest living relatives of plants found in the fossil floras. In fact, tolerances must have evolved through time and TLR varies greatly in different regions.

Physical characters of leaves such as sizes, shapes, and margin types in modern environments allow calibration of environmental factors such as MAT, warm-month and cold-month temperatures, and various parameters of moisture; this is irrespective of taxonomic comparisons of leaves to extant taxa. This calibration, called Climate Leaf Analysis Multivariate Program (CLAMP), is based on >170 samples of modern vegetation. Fossil samples can be subjected to CLAMP, and estimates of various environmental parameters can be obtained.

In CLAMP, which uses Canonical Correspondence Analysis, the relative length of the vectors plotted for each parameter indicates the relative significance of the various environmental parameters. The longest vector is that of calculated moist static energy (enthalpy), which is an environmental factor that is calculated from specific humidity (a measure of actual moisture), MAT, and gravity; these allow calculation of relative altitude at a given latitude. When enthalpy is included in CLAMP, the standard deviations of the residuals in four-dimensional plots indicate that enthalpy is being estimated to a value that corresponds to about 650 m.