Tuesday, August 3, 2010 - 11:10 AM

COS 18-10: Scales of pre-Columbian disturbance around western Amazonian lakes

Crystal N. McMichael1, Mark B. Bush1, and Dolores Piperno2. (1) Florida Institute of Technology, (2) Smithsonian National Museum of Natural History and Smithsonian Tropical Research Institute

Background/Question/Methods Recent evidence from several archaeological sites in Amazonia suggests that forests were extensively modified by humans prior to European arrival (cultural parkland hypothesis). Extrapolating this idea as truth across the basin has major implications for Amazonian ecosystem processes and infers that forests are resilient to disturbance, which could ultimately weaken conservation efforts. Fire, mainly anthropogenic in western Amazonia, is a major landscape modifier, affecting forest composition, succession, and carbon budgets. Pre-Columbian fire and agriculture evidence are also reported in many lake sediment records in Amazonia, but the extent of disturbance around these lakes remains uninvestigated. This study focuses on spatial and temporal fire patterns around three Amazonian lakes (Ayauchi, Gentry, Parker) with known human history to determine whether the ‘cultural parkland' hypothesis applies in these areas. Randomly located soil cores were collected by a hand auger at 20 cm depth intervals, sieved at 0.5 mm, and analyzed for charcoal and phytoliths, which were used to determine forest structure before and after fire events and indicate pre-Columbian agricultural activity. Any charcoal >20 cm depth is assumed to be pre-Columbian aged based on previous literature but charcoal fragments were dated to provide temporal frameworks for fire patterns. Results/Conclusions At lake Ayauchi, Ecuador (N = 10), charcoal > 20 cm was found in eight sites, and maize phytoliths in 6 sites. However, 56% of charcoal fragments (N=9) dated modern, indicating bioturbation affected results. Phytolith analysis show that forest taxa dominated the soils throughout all sites, indicating a very weak disturbance signal even considering bioturbated soils. At lakes Gentry and Parker, Peru (N=77) large charcoal spikes > 20 cm were found in 19.5% of sites, and small charcoal spikes were found at 32% of sites (charcoal present at 52% of sites total). Dated charcoal fragments (100%) (N=7, but more fragments submitted) support the assumption that charcoal > 20 cm is pre-Columbian aged. Phytolith results indicated one site near the lake edge contained maize, and forest taxa again dominated the landscape. Fire and agriculture around these three lakes occurred in pre-Columbian times, but was localized and patchy. Lake sediment records indicated human activity at all lakes; however, no evidence of a ‘cultural parkland' or forest transformations were found around any lake. Lakes are preferred settlement sites because of increased resource availability, therefore random transects in terra firme forests are expected to also show little to no disturbance in most sites.