Question: We were interested if and how variation in frequency and/or size of disturbances affect the dynamics of a montane old-growth forest in Central Europe.

Location: The forest, co-dominated by Fagus sylvatica, Picea abies and Abies alba, is located in Lower Austria and represents one of the few sizable virgin forests in Central Europe.

Methods: We extracted cores from 100 trees using systematic grid sampling (grid cell size 10 m × 10 m) on each of four 1-ha plots distributed across the old-growth remnant of 300 ha. We inferred disturbance events from rapid early growth and release events. For defining release criteria, we applied the boundary line method. We investigated the spatial structure of current age and gap distributions and past disturbance events in grid cells, using a pair density statistic.

Results: The disturbance histories indicate decades with peaks and also extended periods without disturbance. Some peaks occurred synchronously at three of the four plots (1910s, 1930s, 1960s and 1980s). Peaks and gaps in the disturbance chronologies widely agreed with peaks and gaps in the age distributions. Most disturbance events during single decades showed a random spatial distribution.

Conclusions: There is considerable variation in disturbance frequency and/or severity over time. Most disturbance events will rather thin the stand than clear larger areas at once. Following scattered disturbance two pathways occur: (1) gap expansion leading to creation of larger gaps, and (2) gap closure by lateral encroachment or by subcanopy trees growing into the canopy.

Abbreviation: PSP = Permanent sample plot.