GIS analysis of Dakota flora localities.

Abstract

The Cretaceous Dakota Flora from the Western Interior of North America is well known for its angiosperm diversity (435 nominal species). Paleobotanists find this anomalously high level of diversity significant considering the relatively low diversity found in other Cretaceous angiosperm floras, commonly containing 20-25 species. Although some paleobotanists argue that the high diversity is questionable, angiosperm species holotypes from the Dakota can be are still important, if they are restudied and reclassified. Much of the Dakota Flora collection and study can be referred to as historical since it was performed from 1850 to 1920, when little was understood about stratigraphic relations within the Albian to Cenomanian allostratigraphic equivalents to the Kansas Dakota Formation. Further, precise records are not available for geographic locations and stratigraphic positions of Dakota angiosperm fossil collection sites. Some taxonomic questions can be resolved by studying existing specimens from collections worldwide, but unless further fieldwork is conducted, resolution of all angiosperm diversity issues will be difficult. This thesis was undertaken to determine if historical collection sites could be more precisely relocated using geographic information systems (GIS) than using descriptions or using approximations drawn on a paper map, and if GIS analyses could help determine from which stratigraphic horizons the Dakota angiosperm specimens were collected. There were four questions for this research: 1) Can existing location descriptions be used to create a GIS dataset that represents the collection sites as polygons of various shapes? 2) Can GIS analyses be used to refine these collection site polygon locations to smaller polygons that could be more useful for relocating specific historical collection sites? 3) Do spatial analyses of the relationships between the derived collection site polygon locations and other GIS datasets indicate that specimens were collected from a single well-defined geographic area or from a Widespread geographic area? 4) After relocation, do analyses indicate that the collection sites are dispersed within similar enough stratigraphic units to indicate that they were extant contemporaneously? Locations for historical collection sites were defined either as: 1) distances from a known location; 2) directional distances from a known location; or 3) legal descriptions. Legal descriptions were converted into polygon locations, whereas distance and directional distance locations were determined using a combination of buffers and directional indicators. The resulting locations were imprecise. Comparisons with Dakota Formation outcrops helped increase the precision; however, results depended greatly on the quality of collection site descriptions and outcrop location data. Comparisons between the GIS-derived collection site locations and other GIS datasets (surface elevations, Dakota base, Cretaceous base) yielded mixed results. GIS analyses can be useful for planning further field research, but without more and better descriptive locational data or further fieldwork, questions about precise collection site relocation, species diversity, and species distribution cannot be entirely resolved.