DETERMINING TERRESTRIAL BIOME STRUCTURE AND PALEOCLIMATE IN THE NORTHERN NORTH AMERICAN CORDILLERA USING FOSSIL PLANTS: COMPARING COMPARATIVE METHODS IN PALEOBIOGEOGRAPHY

The presence, absence, and abundance of fossil remains in any particular locality have been used, in combination with multiple other coeval assemblages, to recognize patterns of paleobiogeographic proviciality. This approach to understanding geographic regions in the distribution of life forms through deep time has at times provided the foundation for models of paleogeography and paleoclimate on the local to global scale or provided tests of independently constructed earth historical models. The meaning of any paleobiogeographic province with regard to paleoclimatic/paleoceanographic conditions varies, however, depending on the study organisms and the nature of the depositional environments in which they are preserved. Here I review paleobiogeographic methods by which plant fossil occurrences may be used in reconstructing terrestrial biome geography under deep time (pre-Quaternary) climatic and geographic conditions in the nothern North American Cordillera. Such methods include multivariate techniques of presence/absence data, single taxon paleobiogeographic affinity determinations, and coexistence analyses. All methods suffer to varying degrees due to issues involving time-averaging, stratigraphic correlation, and how to reconcile the meaning of taxon absences in an analysis for example. Comparisons of disparate methods of paleobiogeographic analyses of leaf fossil assemblage data from Early Permian (Mystic sequence), Early Jurassic (Talkeetna Volcanic Formation), and Late Cretaceous (Cantwell Formation) strata indicate that using such methods to develop and test models of the accretionary and paleoclimatic history of this region must be done with caution.