Testing predictions of forest succession using long-term measurements: 100yrs of observations in the Oregon Cascades

QuestionMany predictions about forest succession have been based on chronosequences. Are these predictions - at the population, community and ecosystem level - consistent with long-term measurements in permanent plots? LocationPseudotsuga menziesii (Mirb.) Franco dominated forest in western Oregon, US. MethodsOver a 100-yr period, measurements every 5-10yrs of the growth, mortality and regeneration of individually tagged trees in three 0.4-ha forest plots dominated by P.menziesii were used to test predictions derived from chronosequence studies. ResultsPopulation- and community-level predictions generally matched observations: the initial cohort of pioneer species declined exponentially, with the shorter-lived Prunus emarginata (Douglas ex Hook.) Eaton and Arbutus menziesii Pursh disappearing altogether, and long-lived species such as P.menziesii persisting; tree size distribution shifted from a log-normal to a normal distribution, although the establishment of mid- to late-seral species created a bimodal distribution that may represent a transitional phase not usually elaborated in prior work; and mortality shifted from largely density-dependent to increasing amounts of density-independent causes. The observed biomass composition of these forests, even after 154yrs, was still largely dominated by P.menziesii, which was consistent with the prediction from chronosequence studies. The slowing of biomass accumulation as stands aged predicted from ecosystem theory was not consistent with the observation that live biomass accumulated at a relatively constant rate for the 100-yr period. ConclusionPredictions from chronosequences at the population and community level were consistent with long-term observations in permanent plots, whereas those at the ecosystem level were not. At the spatial scale (<2ha) examined with these plots, the high heterogeneity of tree mortality may lead to a multi-modal pattern of net live biomass accumulation with long periods of constant gain interrupted by sudden losses of live biomass.