<p dir="ltr">Succession is a foundational idea in ecology and serves as way to study ecological community assembly. Here, we test the three alternative mechanisms for succession: (I) facilitation, where species arrival order is not random and can be categorized as falling into distinct stages of succession such as early and late species; (ii) tolerance, where species arrival is random, but species interactions that promote coexistence or exclusion determine the community structure through time, and; (iii) inhibition, where species arrival is once again random, but species hold space until they die, and only then can there be a change in species composition through colonization of gaps. To do this, we used a 70-year time series of succession in an oak-hickory forest system in northwest Indiana called the Ross Biological Reserve. Every ten years tree species identity and abundance were measured beginning in 1951. Non-metric multidimensional scaling (NMDS) revealed distinct periods of community assembly that we categorized as distinct early, mid, and late successional communities. We used an index of species co-occurrence (the checkerboard score) and found that early successional communities were largely random, while mid and late successional communities were characterized by nonrandom co-occurrence of species. Alpha and gamma richness increased, and species turnover rate also increased. Specific species relative abundance showed typical patterns of species abundance when classified by their successional status. We conclude that facilitation was the predominant mechanism at the landscape scale, while tolerance was more of a driver on a smaller local scale. We found no evidence of inhibition. Succession at the scale we have documented is difficult to study, because of the long times required to document change. More successional studies of this nature will benefit the field of community assembly and ecology as a whole.</p>