TAPHONOMIC AND (PALEO)ECOLOGICAL IMPLICATIONS OF TROPICAL FOREST LEAF LITTER

Fossil leaf assemblages preserved in fluvial-floodplain deposits represent snapshots that accumulate over ecological timescales and theoretically enable comparison of local species richness and heterogeneity among forests widely separated in space and time. Such comparisons are a primary goal of paleobotany but only become meaningful when they apply a standardized and statistically backed approach to estimating diversity measures traditionally based on individuals from dispersed leaves. The development of a standardized approach to estimating species richness and species heterogeneity based on leaf assemblages is hindered by (1) variable taphonomic bias among depositional sub-environments of fluvial deposits, (2) inconsistent scaling of taphonomic bias across forests with different vegetation structure and source diversity, and (3) a lack of appropriate experimental taphonomic data across a wide range of forest types and depositional environments. Here, we present a quantitative analysis of the relationship between leaf litter samples—analogous to fossil leaf assemblages—and 25 hectares of mapped tropical rainforest in Amacayacu, Colombia, and contextualize these results using a compilation of similar taphonomic studies from forests in Australia, Connecticut, Costa Rica, Maryland, Mexico, Panama, Peru, and Virginia. Similar to previous studies, we find that fluvial sub-environments sample distinct plant communities and that their derivative leaf assemblages may produce widely varying estimates of species richness and species heterogeneity. Autochthonous forest floor samples tend to have lower richness and higher heterogeneity than transported samples derived from the same forest, reflecting differences in spatial averaging during leaf accumulation and real differences in the composition of the source communities. The establishment of guidelines for sampling and comparing ecological diversity measures from fossil assemblages analogous to those included in our taphonomic dataset is a subject of ongoing work.