欢迎访问《生态学杂志》官方网站,今天是 分享到:

生态学杂志

• 研究报告 • 上一篇    下一篇

洞庭湖区水稻土有机氮矿化的模型模拟

李文军1,2*,曾细妹1,彭保发1,杨基峰1,赵迪1   

  1. (1湖南文理学院洞庭湖生态经济区建设与发展湖南省协同创新中心, 湖南常德 415000;2中国科学院南京土壤研究所土壤与农业可持续发展国家重点实验室, 南京 210008)
  • 出版日期:2019-05-10 发布日期:2019-05-10

Modeling of organic nitrogen mineralization in paddy soils in Dongting Lake region of China.

LI Wen-jun1,2*, ZENG Xi-mei1, PENG Bao-fa1, YANG Ji-feng1, ZHAO Di1   

  1. (1Hunan Province Cooperative Innovation Center for the Construction & Development of Dongting Lake Ecological Economic Zone, Hunan University of Arts and Science, Changde 415000, Hunan, China; 2State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing 210008, China).
  • Online:2019-05-10 Published:2019-05-10

摘要: 有机氮矿化是土壤氮循环的关键过程,预测土壤有机氮矿化对于评价土壤供氮能力具有重要意义。本研究利用有效积温、双曲线、一阶(One-pool)和二阶(Two-pool)指数模型拟合洞庭湖区典型水稻土有机氮矿化过程,并分析模型参数与土壤有机氮组分间的关系。结果表明,洞庭湖区不同发育类型水稻土有机氮矿化表现出明显差异,各培养时期累积矿化氮量总体上表现为潴育性水稻土>潜育性水稻土>淹育性水稻土。各矿化模型均可对各试验土壤有机氮矿化过程进行有效拟合,但综合比较模型拟合决定系数(R2)、均方根误差值及参数取值显示,One-pool指数模型拟合效果最差,有效积温模型和双曲线模型次之,Two-pool指数模型拟合效果最优。各模型拟合的不同土壤间有机氮矿化速率常数变异较小,与各有机氮组分均无显著相关性(P>0.05);不同有机氮组分中,氨基酸氮和氨态氮与矿化模型中表征土壤有机氮矿化强度或矿化势的各参数始终具显著正相关关系(r=0.755~0.950, P<0.05)。通径分析结果进一步显示,氨基酸氮是土壤有机氮矿化强度或矿化势变化的最主要决策组分并起直接影响作用。Twopool指数模型将土壤可矿化有机氮分为易矿化氮和难矿化氮两类,相对于其他模型可更为准确地描述研究区水稻土有机氮素的矿化;固定该模型中两类可矿化有机氮库的矿化速率常数取值,有助于提升模型实时定量预测土壤矿化供氮的实用性。

关键词: 地下滴灌, BIOLOG, 基础呼吸, 微生物生物量, 土壤酶

Abstract: Soil organic nitrogen mineralization (SONM) is a critical N cycling process. Quantitative prediction of SONM is essential for assessing soil N supply capacity. In this study, four different mineralization models (i.e., effective accumulated temperature, hyperbolic, One-pool and Two-pool exponential models) were selected to fit SONM process in nine typical paddy soils in Dongting Lake region. The relationships between model parameters and soil organic nitrogen components were analyzed using correlation and path analysis. Results showed that SONM rate substantially varied with soil subtypes. The cumulative mineralized N measured throughout the whole incubation periods generally ranked as Fluvisols > Gleysols > Cambisols. Basing on the coefficients of determination (R2), root meansquare error estimations and model fitted parameter values, all  the four models could effectively simulate the SONM process, of which Twopool exponential model consistently had the best fitting effect, followed by hyperbolic, effective accumulated temperature and the One-pool exponential model. Pearson correlation analysis showed that model fitted SONM rate constants were insignificantly (P>0.05) correlated with soil organic nitrogen components due to their weak variability. On the contrary, all model parameters of characterizing SONM intensity or potential had significant positive correlations with amino acid N and ammonium N (r=0.755-0.950, P<0.05). Furthermore, amino acid N was the primary organic N component directly affecting soil mineralizable N capacity. Our results suggested that the Two-pool exponential model could simulate SONM process more effectively than others as it could reveal the mineralization processes of soil active and slow pools of mineralizable N simultaneously. Thus, Two-pool exponential model could improve the validity and accuracy of SONM prediction by assigning fixed values for the mineralization rate constants of soil active and slow mineralizable organic N pools.

Key words: subsurface drip irrigation, BIOLOG, basal respiration, microbial biomass, soil enzymes.