Countrywide analysis of long-term habitat area changes based on multiple types of historical and contemporary data sources

Combining contemporary and historical data sources in landscape ecological studies help better understand changes in biodiversity loss. Most studies on landscape change apply LC/LU categories as cartographical and remotely sensed data sources support this approach. However, division of land-cover categories into finer categories (e.g. habitat types) offers an opportunity to distinguish natural and anthropogenic processes behind landscape change.
Regarding recent biodiversity loss a broader assessment of natural and secondary habitat types would be necessary (Keith et al. 2009, Tittensor et al. 2014). Habitat specific data could also help increase effectiveness of conservation (Keith et al. 2009) and communicate changes to policy makers and the public.
Habitat-level analysis of long-term changes, however, requires specific methods and historical data sources (cf. Gimmi et al. 2011, Biró et al. 2013, Kaim et al. 2016). For identification of past habitat types we developed a point-based method using iterative habitat identification and information transfer between historical and recent sources. In order to perform it properly it was necessary to have a detailed understanding of the different historical sources (e.g. mapping methodology, legend, potential errors) and the landscape.
We estimated long-term habitat changes for the whole territory of Hungary (93,030 km2). Twenty semi-natural habitat types were analysed between 1783 and 2013 for 7 time periods in 5000 randomly selected sample localities from the Actual Habitat Database (Molnár et al. 2007). We used different sources, as historical maps, archival and recent aerial photos and satellite imagery, botanical and other descriptions, recent thematic layers and field data. ArcGIS 10.1.ESRI software was used for managing the database, maps and exporting frequency and relative frequency data based on estimation probability by Python scripts.
Long-term landscape-scale processes could be divided into trends of natural and secondary habitats. Almost every natural habitat types decreased in extent during the studied 230-year period. Trends of secondary habitats were increasing. Three of those habitats losing more than 90% of their area since 1783 are typical Pannonian habitats having most of their stands inside Hungary (90% area loss is equal to IUCN threshold for Critically Endangered status, Bland et al. 2016).
The method could be applied effectively in many countries to understand long-term habitat trends and recognize targets and effectiveness in habitat conservation and management especially in case of priority habitats of the European Union (Natura 2000). The research has been supported by the project “Sustainable Conservation on Hungarian Natura 2000 Sites” within the framework of the Swiss Contribution Program (SH/4/8) and the GINOP-2.3.2-15-2016-00019 project.