Preservation biases are pervasive in Holocene paleofire records

Abstract

Fire and its controls span several spatial and temporal scales in the Earth System and sedimentary paleofire archives are the primary means of inferring how fire varies on timescales exceeding observational records. However, our understanding of the biases affecting paleofire records remains limited. We address this gap by assembling a dataset of Holocene paleofire records to test whether preservation biases interfere with paleofire interpretations. The dataset contains 40 records composed of a total of 17,225 charcoal accumulation rate (CHAR) samples. We find that the “Sadler effect,” which is the observation that sedimentation rates decrease systematically when measured over longer timescales due to the incorporation of sedimentary hiatuses, is pervasive in these paleofire records. In the compiled dataset, the age ranges of measurement share a negative power law relationship with both accumulation rate (AR; AR = 0.4018*[sample age range]-1.09) and CHAR (CHAR = 1.118*[sample age range]-0.6655), indicating that longer time spans of measurement are more likely to incorporate longer period hiatuses into sediment records. This biases AR measurements, which subsequently bias CHAR values. Indeed, more than half of the paleofire records (n = 21) are composed of CHAR values which share a statistically significant negative relationship with the sample age range of their measurement. To our knowledge, our results are the first to identify this sedimentary bias in Holocene paleofire records. As a solution, we therefore provide an interpretative framework which outlines necessary steps to identify preservation bias in paleofire records and intervals. Lastly, we explore the implications of these findings for paleofire research.