5) Mr. Tom Collier

Restoration of mined lands in hyper-arid environments is challenging and requires spatially explicit indicators of recovery. In drylands, biological soil crusts (biocrusts) are critical to soil stability, water retention, and erosion control, making them valuable indicators of restoration success. This study evaluates the ecological restoration of phosphate mining sites in the Zin mines, Negev Desert, Israel, where topsoil application has been used since 2007. Restored plots of different ages were compared to adjacent natural plots using imaging spectroscopy (400–2500 nm) to detect biocrusts, assess spectral indices, and predict soil biochemical properties. We introduce a novel Biocrust Chlorophyll–Cellulose Index (BCCI) based on absorption features near 670 and 2300 nm and compare it with established indices. Soil organic matter (SOM), proteins, and polysaccharides were predicted from image spectra using a stacked ensemble (PLS-R, random forest, ridge meta-learner), achieving R² of 0.84 (SOM), 0.81 (proteins), and 0.76 (polysaccharides). Among spectral indices, the Brightness Index (BI) and BCCI, along with predicted polysaccharides, best distinguished between restored and natural plots (P < 0.01), reflecting a trend of higher values in older topsoil-restored sites and lower values in sites lacking topsoil. A composite of BCCI, BI, and polysaccharides effectively mapped surface condition, while PCA-weighted indices and Getis–Ord Gi* hotspot analysis identified clusters of biocrust abundance and zones of poor recovery. Together, these tools provide a scalable, spatially explicit framework for diagnosing restoration progress and guiding adaptive management in hyper-arid mining areas, with potential applications to degraded dryland ecosystems globally.