Premium
Reconstructing the spatial distribution of large vegetation formations during the mid‐Holocene in Romania: a predictive modeling approach
Author(s) -
Patriche Cristian Valeriu,
Pîrnău Radu Gabriel,
Roşca Bogdan
Publication year - 2020
Publication title -
journal of quaternary science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.142
H-Index - 94
eISSN - 1099-1417
pISSN - 0267-8179
DOI - 10.1002/jqs.3243
Subject(s) - holocene , vegetation (pathology) , physical geography , steppe , landform , climate change , geology , precipitation , geography , geomorphology , paleontology , archaeology , oceanography , pathology , meteorology , medicine
The present study tests the usefulness of SoLIM software and GIS techniques for the reconstruction of large vegetation formations from the mid‐Holocene period in Romania. We used current reference climate data (temperature, precipitation) and the current extent of ecological regions in Romania to derive climate optimality functions for each vegetation formation. The optimality functions were used to simulate current and past distributions of vegetation. The results showed that the current and mid‐Holocene simulated spatial distributions of vegetation are quite similar. Changes were found for about 14% of the Romanian territory (33 946 km 2 ), the most important being the retreat of coniferous forests in favor of mixed forests and of mixed forests in favor of forest steppe. The former is validated by previous pollen‐based studies showing the transition from coniferous forests to mixed forests that occurred in several areas of the country after the mid‐Holocene. The study demonstrates the potential of this methodological approach to reconstruct past vegetation formations, and at the same time that it is a straightforward and expert knowledge‐based method. Although our application uses only climate factors, the results can be further refined by incorporating additional drivers (soil and landform information, site‐specific pollen and fossil data, wildfire data) for a more accurate inference of paleovegetation.