The late Holocene history of Lake Cahuilla: Two thousand years of repeated fillings within the Salton Trough, Imperial Valley, California

Abstract

To constrain the timing of the past seven lake highstands in the Salton Trough, we compiled 423 radiocarbon dates, of which 284 are reliable and have good stratigraphic control, from paleoseismic and archeological sites in the basin. We developed two OxCal models that assume most charcoal, wood, seeds, and twigs recovered from organic mats at or near the shoreline are derived from material that grew within the lake footprint, and therefore date a dry period between lakes. Charcoal samples collected from lacustrine clastic strata may have also been derived from fires burned during a dry period. As an initial constraint, we assume that samples older than those in earlier lake deposits have age inheritance. Assuming the dates are accurately described by their respective 2σ uncertainties, we ran all dates that would run in a preliminary OxCal model, and then removed those with a poor agreement index as defined in OxCal. From this, of the 423 total dates in the compilation, 151 dates are used in the base model and 149 dates are used in an alternative model, with the differences in the models resulting from choices of whether to include or exclude specific dates that may or may not be representative of a particular dry period between lakes. Where the two models agree, the results are robust, but where the models differ, any differences are taken as uncertainty in the lake ages. Historical accounts and a high-resolution paleohydrologic reconstruction allow us to refine some lake ages.The age windows for the past seven Lake Cahuilla highstands are 1731–1733 CE (Lake A), 1618–1636 CE (Lake B), 1486–1503 CE (Lake C), 1118–1165 or 1192–1241 CE (Lake D), 1007–1070 CE (Lake E), 930–966 CE (Lake F), and 612–5 BCE (Lake G). These ages represent the maximum allowable ranges during which a lake may have filled the basin up to the +13 m highstand elevation; the basin may have been dry for significant portions of each time window, though the lake filling and desiccation episodes may have extended beyond the stated highstand age range for each lake. If the paleohydrologic constraints are ignored, some of the lakes may have initiated earlier, by up to three decades. Additional dates would be needed to further bracket the ages of the earlier lakes. Notably, 120 of the original 284 reliable dates were rejected because they clearly violate stratigraphic ordering, implying that more than 40% of all radiocarbon dates in the Salton Basin exhibit statistically significant age inheritance.