On June 6, 2023, the dam of the Kakhovka Hydroelectric Power Plant was intentionally demolished by Russian troops, who detonated pre-deployed explosives, causing an environmental catastrophe in the Northern Black Sea region. Despite the severe environmental and socioeconomic repercussions, the disaster offered unique opportunities to study vegetation development in the long-flooded but now exposed area. The aim of the present study was to reveal and demonstrate the initial stages of formation of the flora and vegetation in the area of the former Kakhovka Reservoir, combining field research data (flora and vegetation surveys and observations, etc.) and remote sensing techniques (calculation of vegetation indices, supervised machine learning classification, etc.). We visited the area in June and October 2023 and in May 2024; during these visits we georeferenced and outlined various habitat types. These field data were used for geospatial modeling. Our studies found that the number of taxa of vascular plants in the area increased nearly 14 times during the year, thus contradicting initial pessimistic predictions and scenarios of desertification, ruderalization, and large-scale invasions of invasive alien species. We identified main types of newly formed terrestrial habitats, including willow and poplar thickets, marshy habitats, mud-, sand- and shell-covered areas with sparse vegetation. Remote sensing methods, augmented by machine learning techniques, complemented these findings, providing detailed habitat maps for the entire reservoir bed. This study contributes to our understanding of ecological successions and habitat dynamics in post-catastrophic landscapes, informing future management strategies for such areas.
Supplementary Material. Fig. S1 and Tables S1–S3 are available on this website: ukrbotj82-05-488-S1.pdf
Keywords: dam breach, floodplain, habitats, machine learning, restoration, vegetation
Full text: PDF (Eng) 2.43M