Physiognomic and taphonomic studies in New Zealand and Australia : implications for the use of palaeobotany as a tool for palaeoclimate estimation
Measurements of surface uplift rate potentially hold the key to understanding the tectonics of mountain belts and areas of uplift. Wolfe (1993) compiled a multicharacter data set (CLAMP) of the physiognomy of woody dicotyledon leaves with a view to using it to extract climatic information from fossil floras. The limited extent of this data set, along with the anomalous behaviour of some outliers cause me to question the global relationship between physiognomy and climate which has been implicitly assumed in all analyses of the data conducted to date. Additional data collected from native vegetation in New Zealand and Australia are compared to the CLAMP data set. These data include samples along altitudinal transects and from different forest types growing in the same climatic regime. In addition taphonomic samples were collected from lake bottom sediments and their physiognomic signals compared to those of the adjacent living vegetation. The possibility that the relationship between climate and physiognomy is sufficiently non-linear that only local relationships should be sought is investigated. To estimate the climate at a certain flora, resemblance functions are used to select physiognomically similar sites. Estimations of climate are formed using only these sites. The power of this approach to estimate mean annual temperature, mean annual precipitation, mean growing season precipitation, and moist enthalpy is investigated using the modern sites, and varying the number of nearest neighbours and dimensions used as well as the type of ordination. The collection of altitudinal transects has allowed the study of physiognomic change with altitude. Because these transects were collected over a very restricted area it was possible to observe this change without the superimposed effects of changing continentality and variation in latitude. In addition studies made of adjacent floras in similar climatic regimes and taphonomic studies of leaves in lacustrine sediments has allowed the beginning of a realistic assessment of possible errors in climate estimation for fossil sites. Fossil sites examined using CLAMP and related methods are re-examined using the nearest neighbour approach.