湘西稻田土壤–水稻系统硒的生物地球化学特征

Biogeochemical Characteristics of Selenium in the Soil-Rice System of Xiangxi Paddy Fields

摘要: 【目的】研究湘西稻田土壤–水稻系统中硒的分布、吸收、富集及驱动因素，揭示了土壤–水稻硒生物地球化学特征，可为土壤硒资源的合理开发与利用提供依据，对湘西发展特色富硒产品具有重要意义。【方法】通过半方差函数模型、克里金插值分析了湘西稻田土壤–水稻系统中硒的分布，Spearman相关性分析和随机森林模型分析湘西稻田土壤–水稻系统中硒的影响因素。【结果】(1) 湘西稻田土壤硒平均含量为0.415 mg/kg，显著高于中国背景值(0.29 mg/kg)，硒资源整体较为丰富。湘西地区稻米硒平均含量达0.048 mg/kg，56.21%的样本符合富硒标准(≥0.04 mg/kg)，但硒富集系数(BCF均值0.1275)较低，硒转移效率受限。(2) 通过半方差函数模型分析，湘西稻田土壤Se、稻米Se含量与BCF最优半方差函数模型均为指数模型，三者的最优模型R²分别为0.715、0.852、0.789，其块基比分别为0.120、0.106、0.124，空间自相关性极强，变异来源主要受结构性因素影响；根据克里金插值可看出湘西土壤Se主要聚集在湘西东部，稻米Se主要聚集在湘西东南部，基本上都分布在富硒地区，具有聚类特征。(3) 湘西地区石灰岩风化物母质发育土壤Se含量最高，紫色砂页岩风化物母质的土壤硒含量最低，第四纪红色黏土母质区种植的稻米Se含量最高。在本研究构建的随机森林模型中显示，土壤有机质含量是影响土壤Se含量的最重要因素；稻米Se含量主要受气候因素影响，其中日照时长是影响稻米Se含量的主要气候因素，土壤Cu含量是影响稻米Se含量的主要土壤因素。【结论】湘西稻田土壤硒资源总体较为丰富，稻米硒含量达标率较高(56.21%)，其空间分布呈现强结构性自相关(块基比 < 0.125)，土壤硒富集于东部而稻米硒富集于东南部；土壤有机质和气候因子(日照、温度、降水)分别是土壤硒与稻米硒的主要影响因素，硒转移效率(BCF均值0.1275)与土壤Cu、Zn含量呈显著负相关，表明土壤中较高的Cu、Zn含量可能抑制硒向水稻的迁移，是潜在的限制作物硒富集的关键环境因子。

Abstract: [Objective] This study aimed to investigate the distribution, absorption, accumulation, and driving factors of selenium (Se) in the soil-rice system of paddy fields in Xiangxi, revealing the biogeochemical characteristics of Se in soil and rice. The findings provide a basis for the rational development and utilization of soil selenium resources and are significant for the development of distinctive selenium-enriched products in Xiangxi. [Methods] The spatial distribution of Se in the soil-rice system of Xiangxi paddy fields was analyzed using semivariance function models and kriging interpolation. Spearman correlation analysis and random forest models were employed to examine the influencing factors of Se in the soil-rice system of Xiangxi paddy fields. [Results] (1) The average Se content in Xiangxi paddy soils was 0.415 mg/kg, significantly higher than the background value in China (0.29 mg/kg), indicating overall rich selenium resources. The average Se content in rice from Xiangxi was 0.048 mg/kg, with 56.21% of samples meeting the selenium-rich standard (≥0.04 mg/kg). However, the selenium bioaccumulation factor (BCF mean 0.1275) was relatively low, indicating limited Se transfer efficiency. (2) Through semivariance function analysis, the optimal semivariance models for soil Se, rice Se, and BCF were all exponential models, with optimal model R² values of 0.715, 0.852, and 0.789, and nugget-to-sill ratios of 0.120, 0.106, and 0.124, respectively, demonstrating strong spatial autocorrelation, mainly influenced by structural factors. Kriging interpolation indicated that soil Se was mainly concentrated in eastern Xiangxi, while rice Se was mainly concentrated in southeastern Xiangxi, both distributed in selenium-rich areas with clustering characteristics. (3) Soil developed from limestone weathering had the highest Se content, while soil from purple sandstone-shale weathering had the lowest Se content. Rice planted in Quaternary red clay regions had the highest Se content. The random forest model constructed in this study indicated that soil organic matter content was the most important factor affecting soil Se content. Rice Se content was primarily influenced by climatic factors, with sunlight duration being the main climatic factor, while soil copper (Cu) content was the major soil factor affecting rice Se content. [Conclusion] Overall, paddy soils in Xiangxi are relatively rich in Se, and the rate of rice meeting Se content standards is high (56.21%). Spatial distribution exhibits strong structural autocorrelation (nugget-to-sill ratio < 0.125), with soil Se enriched in the east and rice Se enriched in the southeast. Soil organic matter and climatic factors (sunlight, temperature, precipitation) are the main influencing factors for soil and rice Se, respectively. Selenium transfer efficiency (BCF mean 0.1275) is significantly negatively correlated with soil Cu and Zn content, suggesting that higher Cu and Zn concentrations in soil may inhibit Se transfer to rice, representing key environmental factors that potentially limit crop Se enrichment.