Developing a novel topographic-based precision restoration framework for Drylands Ecosystems

Abstract

Ecological restoration in drylands can be enhanced by understanding the composition and spatial distribution patterns of natural ecosystems, shaped by landscape geomorphology, species' microhabitat requirements and resource availability. This study aims to develop a precision restoration framework using UAV based data and statistical models, identify suitable microhabitats for native species in degraded areas based on reference ecosystems, and generate suitability and probability maps to guide species reintroduction. We selected an abandoned semi-arid quarry as a case study and identified five native plant species from a nearby natural reference ecosystem to replicate its ecological conditions. High-resolution orthoimages and Digital Elevation Models (DEMs) were obtained, and topographic attributes were calculated to model species' spatial distribution and topographical suitability. The resulting models were then used to generate suitability and probability maps to apply in a restoration site, revealing that species' spatial distributions are strongly influenced by topographically induced microhabitats, with effects varying among species. The distribution models predicted species presence with AUC values exceeding 0.90, identifying insolation, hillslope position, and runoff-related variables as key drivers of species distribution. This methodology enables more precise and efficient ecological restoration planning in arid zones by optimizing species selection and placement to enhance reintroduction success and survival rates.