Numerous factors can contribute to tree invasions of meadows. Knowledge of the timing and spatial patterning of tree establishment can provide insight into the factors that trigger or accelerate the invasion process. In this reconstruction of tree invasion patterns at Bunchgrass Ridge, we addressed the following questions:

• How have rates of invasion changed over the past two centuries?
• Can temporal trends be explained by climatic variation or do they suggest that other factors are important?
• How is establishment spatially structured?
• Is there evidence of biotic interactions in the spatial associations of trees?
• Have lodgepole pine and grand fir contributed differently to the dynamics of invasion?

We mapped all live (n = 5,486) and dead (n = 1386) trees >1.4 m (4.6 ft) tall in four of the 1-ha (2.5 acre) plots that would be subjected to experimental tree removal. All live trees were aged from basal sections (photo, upper right) or increment cores (photo, lower right).

Age structures were then developed and uni- and bivariate spatial statistics were computed to characterize temporal and spatial patterns of invasion. Based on these age structures, trees were assigned to one of two age classes for the analysis of spatial patterns: young: <90 yr; old: ≥90 yr.