PLANT DIVERSITY OF THE CRETACEOUS-PALEOGENE DECCAN INTERTRAPPEAN BEDS: A COMPARISON OF MACROFOSSIL AND PALYNOLOGICAL DATA

The Deccan Intertrappean Beds (DIB) of India preserve a diverse biota that spanned the Cretaceous-Paleogene (K-Pg) boundary and were paleogeographically isolated, providing data on the response of a unique ecosystem to environmental changes and the role of India in the evolutionary and biogeographic history of many groups. We analyzed six sites with co-occurring palynology and macrofossil data (Ambabagholi (AM), Anjar (AJ), Mahurzari (MZ), Mohgaonkalan (MK), Shibla (SH) and Singpur (SP) from four different DIB regions using ordination and clustering analyses at a variety of taxonomic levels (species, genus, family, order). All are thought to be Maastrichtian in age. A high proportion of records are of unknown affinities within angiosperms. The classic MK locality has significantly more species described than any other locality (n>300, vs. next highest n>40), contributing to a likely sampling size effect. Ferns and angiosperms are represented at all six sites, while gymnosperms, bryophytes, algae, and fungi have been described from only a few localities. Palynological and macrofossil occurrences from the same sites show very little overlap at even the ordinal level. Arecales (palms) are represented by pollen and macrofossils at AJ, MK, and SH, while Alismatales, Ericales, Malpighiales, Malvales, Polypodiales, and Salviniales are in both records at MK. Ferns are present as macrofossils at only MK, but are represented in the palynological record of all six sites. Only ~10% of the families are found in >2/3 of the sites, and at the species level, >95% are singleton taxa. Clustering analysis of the ordinal data highlighted the difference between MK and other sites, and palynological taxon counts clustered more closely with each other than with the macrofossil-based counts from the same locality. No trend was observed that corresponded to the different geographic regions from which samples were sourced. These results highlight the importance of combining palynological and macrofossil data, where possible, to obtain the most accurate view of species diversity, and the need for further scrutiny to distinguish the effects of differential investigator bias from features of ecological, environmental, and/or stratigraphic significance.