Integrating small-scale landscape elements into land use/cover: The impact on landscape metrics’ values

•Line buffers have bigger impact on landscape metrics values.•It is better to use different buffer size that is in compliance with the size of the phenomena in reality.•Integration method does not affect the values of landscape metrics significantly.•Metric's values depend on the vector data model.

Over the last 30 years the use and misuse of landscape metrics has been widely studied. However, there has been less attention on incorporating small-scale landscape elements into landscape analysis. Data type used in the analysis can be either vector or raster, while the raster format is more widely used. However, using large-scale topographical vector databases has several advantages – they cover whole countries with very detailed and accurate topographical data. Despite the high level of detail, their amount in Mb is small, which allows simultaneously to analyse large areas. The peculiarity of vector data is that small-scale landscape elements are mapped as point elements or lines. For calculating landscape metrics, the integration of these features and LULC (land use/cover) polygons is needed. In the current study we investigated how integration of point and linear elements into polygon layers affects the values of landscape metrics. Adding line buffers influenced metrics’ values more than adding point elements. The ensemble of point and linear objects is similar to linear objects. Our study revealed that integrating small-scale landscape elements into land use/cover layers by using buffers gives more realistic values if the buffer size is in compliance with the size of the phenomena in the real world and suitable landscape metrics are chosen. However, the metrics that responded to adding small-scale landscape elements in correspondence with their real world impact on landscape metric values might not always be the best ecological indicators in terms of small-scale landscape elements. Another issue is that values of landscape metrics depend directly on the number of classes determined in the data specification, and on the data model. If the number of mappable point and linear objects changes, or the data model of the linear objects changes, the values of landscape metrics differ.

Graphical abstractFigure optionsDownload full-size imageDownload as PowerPoint slide