PDF(Pubmed)
Abstract：
How to distinguish and quantify past human impacts on vegetation is a significant challenge in paleoecology. Here, we propose a novel method, the error inflection point-discriminant technique. It finds out the inflection points (IPs) of the regression errors of pollen-climate transfer functions using modern pollen spectra from vegetation with different values of the Human Influence Index (HII), which represent the HII threshold values of native/secondary and secondary/artificial vegetation systems. Our results show that the HII value at the native/secondary vegetation IPs is approximately 22 and globally uniform, whereas it varies regionally for the secondary/artificial vegetation IPs. In a case study of the Liangzhu archaeological site in the lower Yangtze River, discriminant functions for pollen spectra from three vegetation types and pollen-climate transfer functions of the native vegetation were established to reconstruct paleovegetation and paleoclimate over the past 6,600 years. Our study demonstrates this method\'s feasibility for quantitatively distinguishing human impacts on paleovegetation and assessing quantitative paleoclimate reconstructions using pollen data.
摘要：
如何区分和量化过去人类对植被的影响是古生态学的重大挑战。这里,我们提出了一种新的方法，误差拐点判别技术。它使用具有不同人类影响指数（HII）值的植被的现代花粉光谱找出花粉-气候传递函数回归误差的拐点（IPs），它们表示原生/次生和次生/人工植被系统的HII阈值。我们的结果表明，原生/次生植被IP的HII值大约为22，并且全球均匀，而次生/人工植被IP的区域变化。以长江下游梁祝考古遗址为例,建立了三种植被类型花粉光谱的判别函数和原生植被的花粉-气候传递函数，以重建过去6600年的古植被和古气候。我们的研究证明了该方法在定量区分人类对古植被的影响以及使用花粉数据评估定量古气候重建方面的可行性。
