Scientific Reports (Oct 2023)

Empirical delineation of the forest-steppe zone is supported by macroclimate

  • Ákos Bede-Fazekas,
  • Péter Török,
  • László Erdős

DOI
https://doi.org/10.1038/s41598-023-44221-4
Journal volume & issue
Vol. 13, no. 1
pp. 1 – 13

Abstract

Read online

Abstract Eurasian forest-steppes form a 9000-km-long transitional zone between temperate forests and steppes, featuring a complex mosaic of herbaceous and woody habitats. Due to its heterogeneity regarding climate, topography and vegetation, the forest-steppe zone has been divided into several regions. However, a continental-scale empirical delineation of the zone and its regions was missing until recently. Finally, a map has been proposed by Erdős et al. based on floristic composition, physiognomy, relief, and climate. By conducting predictive distribution modeling and hierarchical clustering, here we compared this expert delineation with the solely macroclimate-based predictions and clusters. By assessing the discrepancies, we located the areas where refinement of the delineation or the inclusion of non-macroclimatic predictors should be considered. Also, we identified the most important variables for predicting the existence of the Eurasian forest-steppe zone and its regions. The predicted probability of forest-steppe occurrence showed a very high agreement with the expert delineation. The previous delineation of the West Siberia region was confirmed by our results, while that of the Inner Asia region was the one least confirmed by the macroclimate-based model predictions. The appropriate delineation of the Southeast Europe region from the East Europe region should be refined by further research, and splitting the Far East region into a southern and northern subregion should also be considered. The main macroclimatic predictors of the potential distribution of the zone and its regions were potential evapotranspiration (zone and regions), annual mean temperature (regions), precipitation of driest quarter (regions) and precipitation of warmest quarter (zone), but the importance of climatic variables for prediction showed great variability among the fitted predictive distribution models.