engleski

To what extent we can detect secondary succession of grasslands in the dependence of different thematic and spatial resolution of reference data maps?

European temperate grasslands heavily depends on direct (mowing) or indirect (livestock grazing) anthropogenic impact. Without that impact they are exposed to secondary succession process. During that process, they become replaced with shrubs and finally with forests. This inevitably leads to loss of part of biodiversity that we are trying to preserve and protect through the NATURA2000 and similar initiatives. In order to gain information about already lost grasslands, from which we can focus our efforts to most critical or most promising location for grassland conservation, besides current we need historical data. Very often these data differ in their spatial and/or thematic resolution. Here, we compared efficiency of grasslands extinction detection based on medium (Corine Land Cover –CLC) and large (vegetation/habitat) scale maps. Based on the availability of data, we have selected four nature parks in Croatia, two in Continental and two in Mediterranean part. CLC data were available for period 1980–2006. Historical vegetation data originated from 1970s and were made at 1:50000 scale, while current data on vegetation/habitats originated from period 2004–2010 on various scales (1:25000 for parks in the Continental part ; 1:5000 and 1:10000 in the Mediterranean part). To enable comparison between vegetation/habitat maps, we had to reclassify maps to lower thematic resolution (from association to alliance or order). In frame of GIS we have calculated percentage of areas where grasslands disappeared, or are in process of extinction. For large scale maps we had conduct “cell by cell” analyses to detect “extinction phase”, i.e. whether former grasslands are replaced by shrubs or forests. Use of CLC maps for change analyses resulted partly in complete absence of detected changes, while on the same thematic resolution usage of finer spatial resolution map yielded detection of significant changes. Although nature parks differed significantly in “starting” grasslands area (1 and 23.5% for Continental ; 17.3 and 66.3% for Mediterranean), detected decrease was similar at regional level. It was higher in Mediterranean (49.5 and 50%) than in the Continental (38.2 and 30.4%). Higher decrease in the Mediterranean part could be partly addressed to slightly longer timeframe of compared maps. To reliably estimate secondary succession dynamics, data from at least three time periods should be used. Based on our results it seems that spatial resolution is more critical for successful detection of secondary succession in grasslands than thematic one.

protected areas; Croatia; maps; Corine; habitats

nije evidentirano