Peatlands,as unique limno-terrestrial ecosystems,play an important role in water storage,biod-iversity conservation and global carbon cycling.Water table is a key environmental variable controlling the biogeochemical process in peatlands,and hence it is of great significance to reveal water-table changes for fur-ther understanding the evolution of peatlands.Based on one hundred and sixty samples collected from seven peatlands in the Changbai Mountains,Northeast China,in the summer of 2016,2019 and 2021,this study es-tablished a dataset consisting of sixty-four diatom species and eleven environmental variables.The diatom as-semblages registered clear variations along the water-table gradient from 0 to 66 cm.Partial CCA revealed that Depth to the Water Table(DWT)independently explained 8.4％of the total variance in diatom composition,which was much higher than the sole effect of other environmental factors.Diatom-based water-table transfer function was developed using the Weighted Averaging model and the Maximum Likelihood model,respect-ively.Our results showed that tolerance downweighted Weighted Averaging model with classical deshrinking(WA.cla.tol)had the best performance.After removing the samples with residuals＞20％of total measured wa-ter table range(13.2 cm),the WA.cla.tol model yielded the highest coefficient of determination(R2)between diatom-inferred DWT and observed DWT,and the lowest Root Mean Square Error of Prediction(RMSEP)in leave-one-out cross validation.The diatom-based water-table transfer function built in this study provides a new and effective method for quantitative reconstruction of water-table changes in peatlands of the Changbai Mountains.
关键词
硅藻指标/典范对应分析/加权平均模型/水位埋深/定量重建
Key words
diatom proxy/canonical correspondence analysis/weighted averaging model/depth to the water table/quantitative reconstruction
万方数据