A laminated carbonate record of late holocene precipitation/evaporation from Pretty Lake, Lagrange County, Indiana

Abstract

Late Holocene hydroclimate variability in the midcontinental United States is not well understood because there is a lack of high-resolution paleoclimate records constraining historical climate patterns for the region. Here, a 2500-year-long multi proxy lake sediment record from a northern Indiana kettle lake is used to examine spatial and temporal scales of drought and pluvial patterns in the Midwest. Oxygen (18Ocal) isotope analysis of authigenic carbonate and the sedimentary lithic abundance (%lithics) are the primary datasets used to evaluate hydroclimate trends with supporting information from total organic matter, total carbonate and magnetic susceptibility. We additionally derive a record of local evaporation by subtracting the isotopic composition of precipitation (18Oprecip) as characterized by the nearby Martin Lake, IN, record, from the Pretty Lake 18Ocal record. The combined Pretty Lake hydroclimate record documents climate variability during the last 2 millennia and shows that the Midwest has experienced a wide range of evaporative regimes during the late Holocene. We notice a consistent relationship between the Pretty and Martin Lake multi-proxy records; where reduced (increased) evaporative periods and higher (lower) lake levels at Pretty Lake mostly align with increased (decreased) Gulf sourced precipitation and stream erosion with longer (shorter) warm seasons at Martin Lake. Early periods of much drier, and weakened warm-season evaporation patterns dominated from 600 BCE to 900 CE. Evidence of a prolonged period of enhanced warm-season pluvial conditions, with less evaporation and higher lake levels, during the Medieval Climate Anomaly (MCA) between 900 to 1350 CE; and a pronounced century of arid conditions throughout the Little Ice Age (LIA) from 1350 to 1700 CE followed by a gradual decrease in evaporation and rising lake levels starting at 1700 CE and continuing to present. These trends track other Midwest regional hydroclimate climate records, but show an anti-phased relationship with records from the High Plains and western United States regions. This supports the idea that a hydrocilmate dipole exists between the Midwest and western United States driven largely by mean state changes in the Pacific North American teleconnection pattern, but with modification by local and in-lake responses to mean climate states.