Wild cereal species change as an indicator of desertification in arid lands

Soil erosion in arid desertified lands often results in a segmented landscape, each segment with unique geographical and soil patterns, typically dominated by a single cereal species. We hypothesized that this variation reflects changes in soil fertility and desertification state. Two objectives were defined: (i) Determining desertification processes in the land segments of the ravine transition zone using their geographical outlines data and soil properties, and (ii) Determining the possibility of differentiating between wild cereal species using high-resolution satellite imaging. Two challenges arise: long-term cereal species changes and canopy shade differences are minimal, making fieldwork costly. Thus, we used the Migda Ravine in the Northern Negev, Israel, as a case study to evaluate the effectiveness of remote sensing in quantifying cereal cover. Our proposed method, which uses a historical archive of very high-resolution satellite images, alleviates these two problems. PlanetScope® satellites with SuperDove® cameras (3 m pixel⁻¹ resolution) were utilized to study fourteen plots dominated by Stipa capensis (five plots), Aegilops peregrina (three plots), Avena spp.(four plots), and Hordeum spp (two plots). Eight-band images were collected from January 1 to May 31, 2022–2024, to analyze vegetation indices (VI) related to freshness, greenness, and yellowness. The findings indicate that while fresh vegetation reflectance is similar across species, the G-YNDVI, which incorporates the Yellow band (600–620 nm) and Green-I (513–549 nm), can differentiate cereal species from January until May. The integration of satellite imagery, land segment spreading, and soil data can be used to determine desertification or rehabilitation trends. The described methodology and tools can be applied globally to monitor desertification using cereal species coverage data.