Description:
<p>Fine-scale spatial structure is an essential feature of plant populations, controlling pollination, herbivory, pathogen spread, and resource partitioning. Origins of spatial distribution are often obscure in long-established forests, but successional stands offer insight through their physical and compositional simplicity. We tested the hypothesis that spatial structure in forest herb populations arises through a nucleation process in which colonizing species transition from random to clustered distributions through clonal expansion, seed dispersal, and conformity to environmental gradients. Spatial structure was examined in a chronosequence of 40 s growth stands in southeast Ohio, USA. Herbaceous vegetation was recorded in nested plots to describe the evolution of pattern across multiple scales. Spatial distribution was described as the variance: mean ratio of stem number plot-1, and compared between age classes and functional groups. Environmental influence was assessed as the marginal R² value of environmental models predicting stem number. Herb species responded individualistically to stand age and environmental gradients, although all were to some degree clustered across age classes. Dispersal-limited, nonclonal, and annual species were most strongly clustered, suggesting the importance of seed dispersal range and population growth rate in determining spatial structure. Spatial distribution was weakly related to environmental variables. Clustered distributions established early in succession and remained stable for at least 80 years. Pattern formation can be interpreted in terms of nucleation, as we hypothesized, but clusters form earlier than expected. The spatial structure of herb populations in deciduous forests appears to be governed by patterns established during colonization; environmental filtering appears to play a minor role.</p>