STATISTICAL ANALYSES CORROBORATE THE PALYNOSTRATIGRAPHIC ZONATION OF THE LOWER PALEOCENE, DENVER BASIN, COLORADO

A palynostratigraphic zonation of the Paleocene of the Rocky Mountain region established in Wyoming is based on first-appearance datums (FADs) of key species of angiosperm pollen. It has been extended to other basins from New Mexico to Montana and North Dakota. The zonation was applied to the Paleocene of the Denver Basin using samples from the DMNS-USGS Kiowa core. The lower Paleocene Zones P1 through P3 known regionally were identified in the Kiowa core, which is the principal palynostratigraphic reference section for the basin.

Recently a detailed, quantitative study was conducted on samples from the Kiowa core in which relative abundances of all palynomorph taxa were recorded and the data analyzed using a constrained incremental sum of squares clustering technique (CONISS). CONISS defined a series of palynostratigraphic zones that can be compared with Zones P1 through P3 in the Denver Basin. Whereas the zonation based on FADs utilized 10 key pollen species present in 11 samples, the zonation produced by CONISS utilized a database consisting of 124 pollen and spore taxa present in 23 samples. For CONISS, from 300 to more than 1200 specimens were counted per sample.

In an initial run, clustering defined five zone boundaries within the 611-m-thick sampled interval, but lithologies such as sandstone and lignite influenced the composition of assemblages, so samples other than mudstone were omitted and a second CONISS run was conducted. One well-defined zone boundary and two subzone boundaries resulted. The stratigraphic position of the boundary between zones 1 and 2 from CONISS closely corresponds to the position of the Zone P2-Zone P3 boundary based on FADs. A subzone boundary based from CONISS is offset by one sample level from the Zone P1-Zone P2 boundary based on FADs. A lower subzone boundary from CONISS is within Zone P1, suggesting that a subzone not recognized in other Rocky Mountain basins may be present in the Denver Basin, but the data on which the FAD and CONISS zones are based are quite different. The most significant result from the statistical analysis of Denver Basin data is that the two methods for defining palynostratigraphic zones (first appearances of key species and cluster analysis of all taxa) yield closely similar results. Thus, the validity of the “P zones” used in the Denver Basin is corroborated.