ABSTRACT
A pollen analyst seeking long or ancient records from the Amazon basin will be glad to accept any polliniferous deposit that fortune and geography provide: bogs, road cuts, valley fill, old river beds, fossil soil profiles, even archaeological digs. But lakes, particularly those between 100 m and 1000 m across, have advantages that make them the targets of choice. In lakes of this size, the regional pollen rain (the windblown part of the pollen signal) is most reliably represented, with minimal chances of the local overrepresentation that happens in bog or marsh deposits. Pollen preservation is usually excellent in lake sediments. Sedimentation rates are often constant over long periods, thus allowing sedimentation rates to be accurately measured with an adequate array of AMS radiocarbon dates, which in turn allows calculation of pollen influx. Lake sediments are less susceptible to postdepositional disturbance than other deposits. Other paleoecological signals (such as diatoms, sediment chemistry, plant pigments, magnetic intensities, isotope ratios, etc.) are more likely to be valuable in lake deposits than in bogs or soils. And, in the Amazon, lake sediments are the only likely sink for animal dispersed pollen moved from the local watershed in runoff water.
