Principal Component and Time Series Analysis of a 500-year Stalagmite Geochemical Record from Yucatán, Mexico Reveals Climate Variability, Land-use changes, and Volcanic Ashfall

Abstract:

Principal Component Analysis of stalagmite multivariate geochemical records can provide insight into climate variability as well as the frequency of high-magnitude events (i.e. volcanic eruptions) and even land use changes above cave systems. For most environmental proxies, large trace element data sets can pose difficulties for analysis and interpretation due to natural processes acting across wide ranges of time scales and magnitudes with overlapping influences on individual chemical species. To reduce the complexity of geochemical data, we applied Principal Component Analysis (PCA) and Evolutionary Spectral Analysis to a large high-resolution Laser Ablation Inductively Coupled Plasma Mass Spectrometer (LA-ICP-MS) stalagmite trace element data set from northern Yucatán, Mexico (CH-1), from about 1500-2007 CE. In our study, PCA identified five significant principal components (PCs) in this CH-1 record, which explain >83% of the data set’s variability. Our analysis reveals that PC1 responds to overall trace element loading, including both short-lived trace element influxes associated with volcanic eruptions, and sustained land use changes associated with the Spanish settlement and Henequen (succulent plant) production. PC2 reflects prior calcite precipitation associated with regional dry climate anomalies by increasing Sr and Mg substitution in calcite. High loadings for B and Na indicate that PC3 is sensitive to wet climate anomalies. PCs 4 and 5 reflect related but lagged trace element transport mechanisms. Evolutionary spectral analysis results for the PCs reveal the changing influence of solar 11 and 22-year cycles and the 3-7 year El Niño/Southern Oscillation (ENSO) system over the last 500 years. This study adds to growing evidence that speleothems can record multivariate trace element fingerprints of volcanic eruptions, soil erosion, and different styles of climate variability, which can be useful for model verification and sensitivity testing studies.