Validation and limitations of large-scale forest condition indicators – An example from Hungary

Abstract

As the multi-scale study of biodiversity is extremely resource-intensive, proxy indicators taken from various databases are often used to answer questions on a larger spatial scale. Considering differences in the scale and methods, the application of such indicators (especially from different monitoring systems) requires careful consideration and standard methods of validation.In order to demonstrate this, we validated the results of the MAES-HU (National Ecosystem Assessment of Hungary) forest condition assessment (based on the Hungarian NFD – National Forestry Database) with thematically richer, finer-scale field data. We also examined the relationships of some MAES-HU indicators with other variables significant for nature conservation, not currently included in the NFD. We found the MAES-HU scoring was similar to the fine-scale score results in the case of tree species composition indicators, however, less so for structural indicators. The MAES-HU assessment uses the values of the NFD, averaged for forest management units, and thus tends to underestimate structural variety. This highlights a potential loss of important conservation-related information. During the examination of relationships with other indicators that are not included in the large-scale MAES-HU assessment, we found that the presence of large-diameter and old trees correlates with tree-related microhabitats and large standing deadwood, but no relationship was found for other investigated indicators (game pressure, further deadwood indicators). This highlights the need for integrating some key conservation indicators (presence of old and large trees, quantity and quality of standing and lying deadwood) into existing forest monitoring systems in order to optimise the resources dedicated to multipurpose data collection. Our study also highlights that applying indicators as proxies requires the full knowledge of monitoring methods and validated indicator-indicanda relationships.