Multi-aged micro-neighborhood patches challenge the forest cycle model in primeval European beech

As currently framed, the forest cycle model that underlies close-to-nature management in temperate beech forests throughout the globe specifies an or-derly sequence of temporal development within even-aged patches comprising the forest mosaic. Although this model has been widely applied to European beech (Fagus sylvatica L.) forests, the underlying assumptions of disturbance-induced even-agedness (i.e., within-patch age homogeneity) and competition-induced size differentiation (i.e., within-patch size heterogeneity) have not been tested in natural beech forests due to prohibitions on tree coring in primeval forest reserves. In a rare and unprecedented test dataset of spatially explicit tree ages in an old-growth European beech forest, we employed trian-gulated irregular networks of Delaunay triangles to objectively identify natural tree neighborhoods to determine if neighboring (i.e., within-patch) trees were even-or, at most, two-aged. Age differences among neighboring trees (summa-rized in 25-yr age classes) were rarely <25 yrs and mostly >50 yrs, while the few "even-aged" patches were very small (100 m²) and relatively young (<150 yrs). In this first assessment of the assumptions underlying the forest cycle model in European beech, we observed neither the even-aged cohorts ex-pected for disturbance-induced patches in different phases of development, nor the size differentiation among similarly aged trees that should arise from the neighborhood dynamics of competition, self-thinning, and growth. The lack of patches indicating demographic turnover is fundamentally inconsistent with the forest cycle model as it is currently framed. We call for further explo-ration of spatially-explicit tree age datasets to determine the generality of these observations.