6Plant-Best is a novel tool for the selection of the most suitable plant cover against rainfall-induced 7 shallow landslides. It explores the plant-derived likelihood of slope failure reduction under wetting and 8 drying events, respectively. Plant-Best comprises five comprehensive open-source modules built in the 9 freeware R. The modules' objectives range from the spatial detection of landslide-prone zones to the 10 integrated evaluation of plant-derived hydro-mechanical effects on sloped terrain; from the selection of 11 the best performing plant species to the identification of sensitive plant traits. In this paper, we provide 12 a detailed description of the Plant-Best modules and we show how this holistic tool can be effectively 13 employed for plant cover selection in a shallow landslide context. To do so, we demonstrate the 14 application of Plant-Best on a site with a history of slope failures in Northeast Scotland, where the tool 15 is implemented using seven native plant species including both woody and herbaceous vegetation. The 16 results reveal that different plant species were suitable for protection depending on the hydrological 17 conditionsi.e. wetting or drying. Plant effects were limited to the topmost soil and, in general, 18 underweight plants with dense root systems and broad thick canopies offered the best resistance to 19 failure. This suggested that botanically diverse slopes with different plant functional groups are20 desirable for a more effective slope protection. Plant-Best proved to be a relatively simple but robust 21 tool for the detection of landslide-prone zones, the selection and evaluation of plant covers, and the 22 identification of relevant plant traits related to shallow landslides mitigation. The open-source nature of 23 the tool confers a great versatility and applicability to the tool which can be deployed as a multi-24 disciplinary aid to the decision making process. 25 26 27 landscaping, slope protection, R 28 29 30 31 32 33 34 35 109 (i.e. III and IV) to compute pixel-based slope stability under different soil-plant covers and 110 hydrological conditions at user-defined soil depths. The third (Section 2.4) and fourth (Section 2.5) 111 modules generate fixed and stochastic model inputs, respectively. The former generates spatially 112 explicit soil variables through the implementation of a machine-learning algorithm (i.e. Random 113 Forest; Breimar et al., 2002). The latter uses the Monte Carlo method (e.g. Ross, 2006) on readily Eventually, the fifth module (V, Section 2.6) manages uncertainty by calculating a reliability index 116 (Malkawi et al., 2000), performs a series of statistical tests to identify the most suitable plant species, 117 and carries out a sensitivity analysis for the identification of relevant plant traits. 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 2.2. Module I: Landslide-prone zones detector 133 134 This module combines GIS-based path distance and overlay analyses (e.g. Zhu, 2016), and it 135 is envisaged as a first approxima...