Department of Earth System Sciences, Yonsei University, Seoul, Republic of Korea.
Institute for Plant Ecology, Justus Liebig University, Giessen, Germany.
Sci Data. 2024 Sep 19;11(1):1022. doi: 10.1038/s41597-024-03871-3.
Rates of nitrogen transformations support quantitative descriptions and predictive understanding of the complex nitrogen cycle, but measuring these rates is expensive and not readily available to researchers. Here, we compiled a dataset of gross nitrogen transformation rates (GNTR) of mineralization, nitrification, ammonium immobilization, nitrate immobilization, and dissimilatory nitrate reduction to ammonium in terrestrial ecosystems. Data were extracted from 331 studies published from 1984-2022, covering 581 sites. Globally, 1552 observations were appended with standardized soil, vegetation, and climate data (49 variables in total) potentially contributing to the observed variations of GNTR. We used machine learning-based data imputation to fill in partially missing GNTR, which improved statistical relationships between theoretically correlated processes. The dataset is currently the most comprehensive overview of terrestrial ecosystem GNTR and serves as a global synthesis of the extent and variability of GNTR across a wide range of environmental conditions. Future research can utilize the dataset to identify measurement gaps with respect to climate, soil, and ecosystem types, delineate GNTR for certain ecoregions, and help validate process-based models.
氮转化速率支持对复杂氮循环进行定量描述和预测理解，但测量这些速率既昂贵又不能为研究人员所广泛获取。在这里，我们汇编了一个陆地生态系统中矿化、硝化、铵固定、硝酸盐固定和异化硝酸盐还原为铵的总氮转化速率（GNTR）数据集。数据从 1984 年至 2022 年发表的 331 项研究中提取，涵盖了 581 个地点。全球范围内，有 1552 个观测值附有标准化的土壤、植被和气候数据（共 49 个变量），这些数据可能有助于解释 GNTR 的观测变化。我们使用基于机器学习的数据插补方法来填补 GNTR 的部分缺失值，从而改善了理论上相关过程之间的统计关系。该数据集是目前陆地生态系统 GNTR 最全面的概述，也是对各种环境条件下 GNTR 的范围和可变性的全球综合。未来的研究可以利用该数据集来确定在气候、土壤和生态系统类型方面的测量差距，划定特定生态区的 GNTR，并帮助验证基于过程的模型。
