IMPROVED SPACE-FOR-TIME SUBSTITUTION FOR HYPOTHESIS GENERATION - SECONDARY GRASSLANDS WITH DOCUMENTED SITE HISTORY IN SE-HUNGARY

Space-for-time substitution (SFT) is a common method to study vegetati on dynamics; yet, different sire histories make it often unreliable. W e studied the first 38-yr succession on abandoned fields in a Hungaria n steppe landscape, controlling for present and past abiotic condition s, landscape context and disturbance regime. The sites chosen in the l andscape appear suitable for a historically improved SFT. The results of the SFT show that the most characteristic features of the local suc cession are: 1. the very limited colonization of specialist species, t hough this landscape was expected to be a good propagule source for co lonization since there are many species-rich loess grassland parches s cattered in the landscape surrounding the old-fields, 2. the developme nt of a closed, low diversity grassland dominated by perennial general ists, 3. the sharply decreasing rate of succession in 25-38-yr-old fie lds. The old-fields converged in floristic composition with succession . Convergence may be determined by the following factors: 1. low chanc es for woody species colonization (as a consequence of a poor propagul e source for woody species and regular grazing), 2. dominance of Festu ca and other generalists that may prevent further colonization and 3. regular sheep and cattle grazing as a homogenizing factor. Hypotheses were generated based on these results that can be used to plan local l ongterm experiments or observations e.g. to decide whether dispersal l imitation or recruitment limitation determines the low colonization ra te of specialist species observed in this landscape. Since the 25-38-y r-old fields have a very ''characterless'' species composition (and la ck the distinctive features of loess grasslands), future restoration w orks will need to focus on the artificial diversification of these sec ondary grasslands, ii the goal is a species-rich grassland that resemb les the region's ancient loess grasslands. Improved space-for-time sub stitution seems to be a useful tool to study local vegetation dynamics . Detailed historical reconstruction could improve the reliability of SFT studies; unfortunately, reconstruction of the history is not an ea sy task and the site choice is never perfect. The challenge of histori cal reconstruction may be one reason why SFT studies of sires with doc umented history are rare. Improved space-for-time substitution may als o be an appropriate tool to design restoration projects.