Abstract: Aims Space-for-time substitution (SFT) is often used for vegetation status estimation during the recovery process of deserts. However, the evaluated accuracy of SFT remains uncertain. An eight-year located observation was used to assess the validity of SFT for vegetation state prediction.
Methods This study analyzed a chronosequence of Caragana microphylla Lam. plantings using the located observation method to test the accuracy of SFT for vegetation state prediction in the mobile sand dunes of the Horqin Sandy Land in northeastern China from July 2005 to June 2013.
Important findings According to SFT, simple vegetation parameters (density, coverage and biomass) were found to be unstable, while sophisticated vegetation parameters (species diversity and evenness) were relatively stable across the experimental treatments during the study period. Conversely, both the simple and sophisticated parameters were found to be relatively stable when tested using the located observation method. Furthermore, most simple vegetation parameters slightly increased, while sophisticated parameters slightly decreased after eight years of field observations. Thus, long-term restoration management facilitated improvements in the simple parameters, but may have adversely impacted the sophisticated parameters in the post-restoration community. Our results suggest that sophisticated vegetation parameter states can be predicted by SFT, while simple vegetation parameter states are not well predicted by SFT. In conclusion, located observations or other effective evaluation methods must be employed to offset the deficiency of the SFT method for the prediction of vegetation parameters.

Key words: community structure, desertification, location observation, restoration, stability