Development of a multiplicative model to explore how biological nitrogen fixation responses to multiple edaphic factors in paddy soils.

Various sole edaphic factor has been screened as controlling factor for paddy biological nitrogen fixation (BNF), however, no single factor can interpret different results from different experiments. Paddy BNF might be influenced by several factors simultaneously at various degrees, but we still lack a model to predict BNF as a function of multiple edaphic factors. Developing this model will help us to estimate BNF based on soil properties and improve policy decisions to enhance paddy BNF. In this paper, 27 paddy soil samples with different soil properties were collected from the southwest to the northeast in China. These soils with rice plantations were incubated for 77 days in an airtight, transparent, and ¹⁵N₂ enriched growth chamber to measure BNF. Results showed that BNF amounts ranged from 0.66 to 12.3 kg ha^-1 and had a lognormal distribution. A multiplicative model was developed, the R code and raw data were deposited in this project. This multiplicative model could explain 86% variance of the BNF in various soils. In each site, the controlling factors were screened and corresponding measures to enhance BNF were proposed. Our study provided a systematic framework on how BNF responded to a range of edaphic factors simultaneously in paddy soils, and this work might also give inspiration for the study of other ecological functions.

诸多单一土壤因子（edaphic factor）曾被筛选为水稻田生物固氮（biological nitrogen fixation, BNF）的调控因子，但尚无单一因子能够解释不同试验间的结果差异。水稻田生物固氮可能受多种因子以不同程度共同调控，但目前仍缺乏以多土壤因子为变量的生物固氮预测模型。构建该模型将有助于基于土壤属性估算生物固氮量，并为制定提升水稻田生物固氮的政策提供依据。本研究从中国西南至东北区域采集了27份具有不同土壤属性的水稻田土样，将这些种植过水稻的土样置于密闭、透明且富¹⁵N₂的生长箱中培养77天，以测定其生物固氮量。结果显示，生物固氮量介于0.66至12.3 kg·ha⁻¹之间，且符合对数正态分布。本研究构建了乘性模型，相关R代码与原始数据已上传至本项目。该乘性模型能够解释不同土壤中生物固氮量86%的变异。针对每个采样位点，本研究筛选了其生物固氮的调控因子，并提出了相应的提升生物固氮的措施。本研究构建了水稻田生物固氮对多土壤因子协同响应的系统性分析框架，该工作也可为其他生态功能相关研究提供参考思路。