Quantification of uncertainties in fossil leaf paleoaltimetry: Does leaf size matter?

The utility of multivariate foliar physiognomy, specifically the Climate Leaf Analysis Multivariate Program (CLAMP), to yield reliable estimates of enthalpy and, hence, paleoelevation has been demonstrated by comparison with other proxies, yet concerns have arisen regarding uncertainties arising from (1) apparent ambiguities in the scoring regime and (2) the way leaf size is scored. Regarding the first concern, scoring ambiguities are examined by reporting on scoring tests with novice users and interlaboratory comparisons. The uncertainties were found to be less than those arising from the statistical methodology underpinning CLAMP. In respect to the second concern, the effect of removing all size data both from modern test sites and fossil data was tested. Specifically, the effect of removing leaf size data from the 15 Ma Namling data set from south central Tibet was investigated. Removal of all size data from modern sites demonstrated that size data contributes little to estimates of mean annual temperature and enthalpy. Similarly, the removal of leaf size information from the Namling data set alone, but with calibration unchanged, and from both the Namling site and calibration sites, this time with recalibration, still yield paleoelevation estimates that have been independently matched by oxygen isotope techniques. Moreover, the removal of all leaf size information results in only small increases in uncertainty (±52 m).