镉
追踪
环境科学
光谱特性
光谱分析
污染物
压力(语言学)
生物系统
水稻
污染
农学
土工试验
作物
土壤科学
光谱特征
土壤污染
蒙特卡罗方法
光学传感
跟踪(教育)
遥感
环境压力
电厂系统
土壤水分
光污染
环境化学
镉暴露
表征(材料科学)
作者
Yang Yuping,Nan Wang,Qinglin Li,Ruiwu Xu,Yong He,Yongpeng Zhao,Zhijun Wu,Hui Huang,Lijia Xu
标识
DOI:10.1021/acs.est.6c10708
摘要
Abstract Plants exposed to heavy-metal-contaminated soil are subjected to pollutant pressure, making timely monitoring of plant condition important for assessing their responses to soil pollution. Although spectral sensing is widely used for nondestructive monitoring of plant stress, the cross-domain association between plant optical signals and Cd-induced physiological responses remains unclear. Here, we integrated spectral sensing with deep learning, Monte Carlo simulation, and omics analysis to systematically explore how spectral signals and physiological responses jointly support the monitoring of Cd-induced stress in rice stems. We found that Cd exposure reorganized interface-associated stress, protective chemistry, and redox regulation into coordinated Defense, Repair, and Adjustment signatures. These biological changes were differentially expressed across the visible, red-edge, and near-infrared spectral regions, producing a reweighting of absorption- and scattering-sensitive information. The agreement between spectral variation and biological stress signatures further established a cross-domain link between stress-state formation and optical detection. Benefiting from this mechanism, we achieved high-accuracy nondestructive detection of rice stem Cd content and noninvasive characterization of Cd-induced responses. A portable spectrometer–smartphone implementation further demonstrated field-oriented feasibility. These findings provide a biological and optical basis for tracking stress progression and supporting timely, feedback-guided crop management.
科研通智能强力驱动
Strongly Powered by AbleSci AI