Katarzyna Bzdęga1, Adrian Zarychta1, Barbara Fojcik1, Roksana Zarychta2, Alina Urbisz1,
Sylwia Szporak-Wasilewska3 and Barbara Tokarska-Guzik1
1University of Silesia, Faculty of Biology and Environmental Protection, Department of Botany and Nature Protection, Jagiellonska 28, 40-032 Katowice, Poland
2Pedagogical University of Cracow, Institute of Geography, Department of Geoinformation and Environmental Research, Podhorążych 2, 30-084 Cracow, Poland
3Warsaw University of Life Sciences, Faculty of Civil and Environmental Engineering, Water Centre Laboratory, Nowoursynowska 166, 02-787 Warsaw, Poland
Distribution modelling focused on forecasting the spread of plant species is an important and desired method that supports resource management, conservation decision making, as well as monitoring and control of invasive species. The goal of this study was to model the current spatial distribution of invasive plant species and to predict their coverage using geostatistical tools. The main species of interests were: Heracleum sosnowskyi and Fallopia spp. (Fallopia japonica and F. × bohemica). The field data were acquired during days of species optimum development in selected areas during 2017 and 2018 to generate spatio-temporal cover patterns of the invasive plant species. We then predicted their spatial distribution in 2019 with the use of Sequential Gaussian Simulation (SGS). To visualise the spatial variability of the analysed species, spherical variograms in both areas and years were created. In order to verify the correctness of the conducted modelling, the cross-validation was tested. The values of each calculated error (e.g. mean error, root mean square, average standard error, mean square error) oscillated from -0.007 to 0.883, whereas root mean square standardized error took values more than one (i.e. between 1.067 and 1.454). These error values indicate correctness of SGS. For the purpose of this study, the collected field data were tested for the presence of spatial autocorrelation by computation of the Moran’s I and z-score for the analysed species on these areas in both years. The results showed that the Moran’s I Index values (i.e. 0.521–1.076) and z-score (i.e. 7.683–13.412) indicate significant clustering of spatial phenomena, which confirmed the existence of spatio-temporal patterns (Fig.1). Although further research is needed using other species, the study results indicate that these determined spatio-temporal patterns with calculated autocorrelation may be used to better predict the spread of invasive species in subsequent yearsand will therefore enhance the effectiveness of management strategies.