The spatio-temporal pattern of release signals and tree growth in Fagus-Abies-Picea old-growth forests reveals unsteady gap-phase dynamics

Abstract

Disturbances have been recognized as a key factor shaping the species composition, structure and dynamics of natural forest ecosystems. In Europe, where forests driven by spontaneous processes have survived in relic form, knowledge about natural disturbance regimes is still fragmentary. To expand this knowledge, we reconstructed stand-level growth and analyzed the spatio-temporal pattern of release signals in the increment chronologies of individual trees as indicators of disturbance events in the Western Carpathians (Central Europe). The study was carried out in five old-growth forests formed by Fagus sylvatica L., Abies alba Mill. and Picea abies (L.) H. Karst. Depending on the stand, the analyses included tree-ring series of 84–193 trees sampled over areas of 5.9–13.6 ha and aimed at determining (1) the spatio-temporal pattern of disturbance severity over the last two centuries, (2) whether disturbances have been synchronized in time across the study sites and (3) whether disturbances have induced pulsed dynamics of stand development manifested as fluctuations in radial tree increment at the level of entire stands. In the period 1850–2010, the percentage of decades with the proportion of released trees < 10, 10–20, 20–30 or more than 30% was 38, 41, 14 and 7%, respectively, and no instances of severe disturbances simultaneously impacting an extensive area and releasing the vast majority of trees were found. The release events were only weakly synchronized at the between-stand level. The spatial distribution of released trees varied over the decades, with a shift toward spatial independence for the most severe disturbances. At the stand level, the interchanging periods of increasing/decreasing tree growth lasted between 24 and 36 years, with the exception of one stand in which this period lasted 54 years. The revealed fluctuations in tree growth attributable to changes in stand density were relatively small and accounted on average for 7% of the total variation in annual tree increments. This suggests that local level disturbances introduce structural heterogeneity and strongly modify tree growth, but at the stand level, their effect is dispersed and causes only minor fluctuations. An over-dispersion of decadal release frequencies compared to the random model and spatial correlation of disturbing events on the one hand, and the lack of extensive disturbances, frequent occurrence of multiple releases in tree life histories, and small fluctuations in the reconstructed growth at the stand level on the other hand, suggest a disturbance regime which goes beyond random processes in a strict sense and is thus not entirely compatible with the classical model of gap-phase stand dynamics.