FROM SAMPLES TO SENSORS: INTEGRATING FIELD, ANALYTICAL, AND SPECTRAL TOOLS FOR SUSTAINABLE PHYTOREMEDIATION IN ARID LANDSCAPES

Abstract

Revegetation of mine-impacted landscapes in drylands remains a major environmental challenge due to extreme substrate conditions, water limitation, and drought. To advance understanding of vegetation establishment and monitoring in these environments, this dissertation integrates ecological field surveys, laboratory-based elemental analysis, and spectral sensing approaches across multiple spatial and analytical scales.First, one of the few successfully revegetated Cu–Mo tailings ponds in the semi-arid U.S. Southwest was investigated to identify key drivers of plant establishment and elemental accumulation. Hierarchical clustering revealed that vegetation communities at different successional stages corresponded to specific substrate properties. Several native species exhibited elevated uptake of Cu, Se, and Re, verified through X-ray fluorescence screening of herbarium specimens, establishing robust benchmarks for identifying new (hyper)accumulating species. Second, to develop efficient analytical tools for assessing plant elemental composition, a portable monochromatic X-ray fluorescence analyzer (MXRF) was validated using plant-based NIST Standard Reference Materials and field samples. New performance guidelines were established for sample preparation, analytical precision, recovery, and detection limits, showing that optimized MXRF methods provide reliable quantification for Cu, Mn, Fe, and Zn. These results offer the first comprehensive quality standards for monochromatic pXRF in environmental and phytomining applications. Finally, we explored hyperspectral spectroradiometry as a non-destructive method for detecting plant metal(loid) accumulation and physiological stress. Controlled hydroponic experiments with the Zn-hyperaccumulator Arabidopsis halleri revealed phase-specific physiological and spectral responses to sequential Zn exposure. Distinct Zn-response phases reflected stages of adjustment, stagnation, and acclimation, with the red-edge region (≈686–750 nm) consistently emerging as the most informative for differentiating Zn-induced changes. Collectively, this research establishes an integrative framework linking soil–plant interactions, elemental analysis, and spectral signatures to monitor vegetation recovery and metal(loid) dynamics on legacy mine sites. The combined methodological advances in X-ray fluorescence and hyperspectral sensing contribute toward the development of rapid, non-destructive tools for evaluating ecosystem rehabilitation and supporting future phytomining strategies in dryland environments.