Abstract

The rodent midden database from the Atacama Desert has been more than 15 years in the making and with over 1500 middens collected and more than 600 dated with radiocarbon, now rivals the North American packrat midden record. In this talk, we review the progress and results accomplished with this unique archive. Major progress has been achieved in relating different midden proxies to each other and in quantification of environmental and climate variables in this extreme environment. These include advances in relating midden floras (pollen, macros and cuticles) to climatic gradients and precipitation seasonality (the Atacama straddles the winter and summer rainfall regimes) and the development of transfer functions that relate, for example local pollen rain or plant species present as macrofossils in modern middens to current gradients across more than 10° in latitude and over 3000 m in elevation. New proxies have also been developed, such as the use of Abrocoma (Abrocomidae) fecal pellet diameters to estimate past changes in productivity, which in the hyperarid Atacama Desert is directly related to changes in rainfall. Other advances include major contributions to understanding past biogeographic changes of the Atacama flora and the consequences that past climate events have had in creating both refugial and relict plant populations in the central Andes. These studies have also confirmed other major known patterns, such as the likely isolation of the highly endemic coastal Lomas floras throughout the late Quaternary. Ongoing studies are now relating the history of Polylepis tarapacana (the tree found at the highest elevations in the world) to paleoenvironmental changes and archeological dynamics, as well as study of rodent middens from archaeological sites used to reconstruct local environments and resource use. New frontiers now include the use of stable isotopes to reveal past changes in the N and C cycles in the Atacama and we are now exploring the use of δD on leaf waxes preserved in rodent pellets to reveal changing sources in precipitation. Future research will include gap analyses (spatial and temporal) of the current database and using midden floras to provide historical evidence for the rules of assembly and migrational histories of Atacama Desert plant communities. More midden research is also needed for the Altiplano, where many plant species that occurred at lower elevations in the late Pleistocene are now present in this high altitude habitat in what appear to be refugial populations. Such efforts, however, will be heavily dependent on the development of extensive databases to manage the complex paleoecological information that the midden record is now capable of providing.