Search

Questions: What are the characteristics of the spatial patterns of soil seed banks (SSBs) at local scales? How does the spatial distribution of SSBs relate to above‐ground vegetation? Location: The piedmont plain of the eastern slope of Gandur mountain, Wuhai, northwestern China. Methods: A plot of 50 m × 50 m square, divided into 100 quadrats (5 m × 5 m) and 100 subquadrats (1 m × 1 m), was prepared to conduct a field survey of vegetation and SSB sampling. We analyzed the spatial pattern of 196 samples at different scales and directions with geostatistical analysis. The relationship between above‐ground vegetation and SSB was quantified by integrating geostatistics, point pattern analysis, and spatial comparison (IGPS). Results: The SSB density in the study area was 404.5 ± 352 seeds·m−2. SSBs of shrub species and herbaceous species Ⅱ (occurring only below‐ground) had a moderate spatial autocorrelation, and their spatial dependence increased with increasing scale. The SSB of herbaceous species Ⅰ (occurring both above‐ground and below‐ground) had great spatial autocorrelation, and its spatial dependence showed an initial decrease and then increased. Furthermore, SSBs showed lower spatial heterogeneity in the prevailing wind direction than in other directions. The spatial patterns of SSBs of shrub species and herbaceous species Ⅰ correlated with shrubs within a specific range, but not with herbs, while the SSB of herbaceous species Ⅱ did not have a spatial relationship with above‐ground vegetation. Conclusion: The spatial pattern of SSBs had scale dependence features, which exhibited clustered distribution at a certain scale, and were significantly correlated with above‐ground shrubs. Therefore, we suggest subsampling larger quadrats with smaller sample units near shrubs in desert areas. Our workflow seems efficient for determining the spatial pattern of SSBs at medium small scale in arid areas. [ABSTRACT FROM AUTHOR]