Classification and dynamics of developmental stages in late-successional temperate forests

Late-successional forests often have complex disturbance histories that can result in stands with widely varying structure, ranging from young pole stands to uneven-aged old growth. Arranging stands in chronosequences, however, is problematic because 'stand age' is not a meaningful concept for multi-aged stands and 'time since last stand-replacing disturbance' often cannot be determined from tree-ring evidence. In this paper, we describe a systematic approach for classifying developmental stages in late-successional forests using structural metrics known to be correlated with key ecological properties such as total biomass, carbon storage, stand production rates, and wildlife habitat. While conceptually based on the amount of aggregate crown area occupied by different size classes of trees, the computations in this study, for ease of use, are based entirely on absolute and relative basal area of four size classes (saplings, poles, mature, and large trees). Eight forest structural stages are recognized, including four stages of old growth (early-, mid-, late-transition, and steady state). The method was used to classify developmental stages of 70 primary northern hardwood stands (Acer-Betula-Tsuga) in large landscape reserves in upper Michigan, USA. The degree to which the developmental stages mimic underlying temporal trends in stand dynamics was investigated with the aid of 30-year permanent plot records in primary forests and multi-century simulations using the CANOPY forest dynamics model. Results indicated good correspondence between the postulated developmental sequence in the 70 field stands compared with CANOPY simulations of structural changes over time and changes observed on the permanent plots. Results support the Bormann-Likens hypothesis that the number of large trees reaches a maximum toward the end of the lifespan of an even-aged cohort and subsequently declines in the steady state. While most of the field plots were uneven-aged, both simulations and field data suggest that developmental pathways of even-aged and multi-aged stands after disturbance are very similar and are difficult to distinguish based on the form of the size distributions. In the study area landscapes, 78% of the stands were classified as old growth, with 39% in early or mid-transition and 39% in late transition or steady state. The method should be useful, with appropriate site-specific modifications, in evaluating the degree to which stands meet structural goals under ecological forestry methods, in monitoring responses to environmental change, and in examining biotic changes along a gradient of structural development following disturbance.