THE IMPACT OF CLIMATE AND PRE-COLUMBIAN AGRICULTURE ON SOIL AND NUTRIENTS IN SOUTHWESTERN COSTA RICA

Lake sediments are useful archives for examining local land use and erosion over time. In particular, the amounts and types of phosphorus (P) eroded from soils into lake sediments may be indicators of nutrient cycling and bioavailability as a function of soil development. This is important on a broader scale because weathering of continental material is the only source of new P into the global P cycle.

Human impacts on landscapes have been recorded on many temporal and spatial scales. An example of this is found in Laguna Zoncho, a lake in southwestern Costa Rica, where a lake sediment core reveals clear changes from forest to maize agricultural ecosystems over the past 3000 years. To examine the effects of soil disturbance and nutrient dynamics in the region, we carried out detailed analysis of P fractions related to soil development on sediment samples from a Zoncho core. These P fractions, termed mineral P, occluded P (iron bound), and organic P change as a function of increased soil maturation, with mineral P dominating in immature soils and being replaced by occluded and organic P fractions as soils mature.

Results from Laguna Zoncho reveal two interesting features of terrestrial nutrient cycling and soil development in this region. Despite pollen evidence showing clear and profound shifts in plant species related to maize agriculture, the net soil nutrient status has remained constant through time. The composition of P is much more organic rich than in other temperate settings studied by us previously, likely due to high rainfall and temperatures in the region. Throughout the ~3000 yr record, a slight shift toward higher % organic P exists, which might be the result of increased dissolved nutrient input into the lake and higher in-situ algal production. Additionally, a strong transient shift toward higher occluded P occurs around ~400 years ago, and coincides with the re-advance of native forests. This could be due either to increased erosion, effectively mining down to deeper soil layers with higher occluded P, or to a short-term (i.e. ~100 year) interval of arid conditions resulting in more occluded P on the landscape. These two, apparently disparate hypotheses will be tested by examining regional paleo rainfall records, particle size analyses in Laguna Zoncho sediments, and pollen records from this and other sites.