The aim of the study was to determine the burnt areas, response of the floodplain ecosystems to burning, and to determine the difference between them and the control sites using the remote sensing data. The impact of burning was assessed using the spectral indices: Normalized Difference Vegetation Index (NDVI), Green Chlorophyll Index (GCI), Specific Leaf Area Vegetation Index (SLAVI), Normalized Difference Moisture Index (NDMI) and Moisture Stress Index (MSI). It was established that during the winter/spring period of 2020, fires in the territory of the Lower Dniester NNP covered almost 4453 hectares (33.1% of the floodplain ecosystems area). Simulation showed that the difference between the burned and control areas gradually decreased over the vegetation period. Maximal difference was observed at the beginning of the growing season (the quality of group selection was 43.2%), which indicates forming of relatively individual environmental conditions for the floodplain vegetation development. In spring, the difference was equal to 20.3%, due to development of <i>Phragmites australis</i>, the edifier of the plant complexes of wetland ecosystems, whose stems have developed a leaf blade. The summer period was characterized by further decrease in the group selection quality (19.9%), owing to equalization of environmental conditions for vegetation development. The autumn period was characterized by dry conditions and end of the growing season, which also contributes to the almost complete disappearance of differences between the plots (14.5%). Based on the analysis of similarity (ANOSIM), it was found that degree of difference between the burned plots was small, owing to local conditions and burning intensity, and amount of dry organic matter in certain areas. It was found that gradual decrease in differences during the vegetation season took place only at winter (February) burning. The influence of spring (March, April) burning resulted in significant variability of the similarity degree over the vegetation period.
Read full abstract