Regional features of pollen R-values in China

Abstract

Establishing a quantitative relationship between modern pollen and vegetation remains a pivotal but intricate pursuit within the field of Quaternary palynology. The R-value, a well-established and traditional metric characterized by its simplicity and direct applicability, serves to calibrate the nonlinear disparities between surface pollen and modern vegetation. This study entails the construction of a comprehensive pollen R-value dataset for China derived from R-values published between 1987 and 2021. The dataset, compiled after rigorous screening, calibration, and standardization processes, encompasses 898 modern pollen sampling sites and 2115 pollen R-value data entries, encapsulating 152 pollen taxa spanning 65 families and 113 genera. Synthesizing the extracted R-values yielded the following insights: (1) A macrospatial analysis validates previously established knowledge at a site-specific scale. Evidently, pollen R-value variations across China are statistically significant. Approximately two-thirds of pollen taxa exhibit underrepresentation (e.g., Taxodiaceae, Corylus, Nitraria, Tamarix, Cyperaceae, Poaceae, and Fabaceae), while the remaining one-third display overrepresentation (e.g., Pinus, Picea, Betula, Ephedra, Chenopodiaceae, and Artemisia). The degree of underrepresentation surpasses that of over-representation, and the representation patterns of Castanea, Quercus, Polygonaceae, and Asteraceae are contingent upon vegetation types. (2) Pollen R-values follow distinct spatial distribution patterns within China. In the woody vegetation region of eastern China, R-values progressively decline from north to south, correlating with rising temperatures and precipitation. Conversely, in herbaceous vegetation regions of northern and western China, R-values decrease from east to west and from northeast to southwest, corresponding to increased aridity. Nevertheless, pollen R-values manifest variation contingent on pollen taxa, vegetation types, and climatic regions, even differing for the same taxa across varying vegetation types and climatic conditions. This highlights the intricate nature of pollen R-values and their interpretation of pollen-vegetation relationships. (3) Pollen R-values and relative pollen production estimates exhibit resemblances and a modest positive correlation. However, adjudicating between them as representatives of vegetation requires nuanced consideration, as both metrics convey pollen representation within vegetation, demonstrating the multifaceted relationships they share with modern vegetation. Further recommendations suggest that when assessing pollen representation in modern vegetation, fossil pollen content should be weighted using either the median or log-transformed R-value. This approach underscores the necessity of comprehensively accounting for divergences and convergences across various spatial scales and vegetation types, particularly the disparities observed within identical pollen taxa across dissimilar regions.