A secondary succession sequence of 14 northern hardwoods stands was sampled for forest floor organic matter and nutrient content. During the first 15 yr following clear cutting, the forest floor decreased by 30.7 Mg/ha, a decline of over 50%. The decrease in the forest floor and slash (logging residue) may be greater than the increase in the living biomass. During the next 50 yr the forest floor increased by 28.0 Mg/ha and by year 64 was within 5% of an asymptote of 56.0 Mg/ha. Nutrients were analyzed in 6 of the 14 stands. Magnesium, potassium, and nitrogen concentrations showed no successional pattern. However, calcium concentrations were significantly higher in the stands in which forest floor mass was low. The initial decrease in forest floor mass is attributed to lower leaf and wood litter fall and to more rapid decay resulting from higher temperature, moisture content, and nutrient levels and to early successional litter being more easily decomposed. The recovery of the forest floor is explained primarily as resulting from the rapid increase in the quantity and diameter of wood litter fall. JABOWA, the northern hardwood forest growth simulator, predicts a maximum rate of increase in woody litter by years 10—20 with a leveling off by years 30—50. An apparent asynchrony in function of the forest floor and slash as nutrient sources may be important to the recovery process. During the first 15 yr the forest floor is a major source of nutrients, releasing a net amount of approximately 800 kg/ha of nitrogen. During this period nitrogen immobilization in the decay of slash may account for as much as one—half of the nitrogen released from the forest floor. After year 15 the forest floor is no longer a source but a sink for nutrients as nutrients and organic matter accumulate. By year 15 the slash probably shifts in function from a sink to a source, providing nitrogen for the continuing rapid nitrogen accumulation in vegetation beyond year 15.

The fire disturbance regime and forest structure prior to Euro‐American settlement (AD 1883) of a southwestern ponderosa pine (Pinus ponderosa) landscape were quantified in order to establish reference conditions as a baseline for ecosystem management. The mean presettlement fire interval between 1637 and 1883 was 3.7 yr for all fires and 6.5 yr for widespread fires, but fire has been excluded from the study area since 1883. Forest density increased under fire exclusion from an average of 148 trees/ha in 1883 (65 pines, 80 oaks, three other species), an open forest dominated by relatively large ponderosa pines, to 1265 trees/ha in 1994/1995 (720 pines, 471 oaks, 74 others), a dense forest characterized by relatively small and young trees. Species composition has shifted toward greater dominance by Gambel oak (Quercus gambelii) and conifers less adapted to frequent fires: white fir (Abies concolor) and Douglas‐fir (Pseudotsuga menziesii). The reference presettlement conditions can be applied to management of this ecosystem in two ways. First, reference conditions are a benchmark against which to evaluate contemporary conditions and future alternatives. The comparison shows that the contemporary forest is well above the range of presettlement variability in forest density, and both live and dead fuel structures have developed that can support high‐intensity wildfire. Second, reference conditions can serve as a goal for ecological restoration treatments.

Ecological restoration is the process of reestablishing the structure and function of native ecosystems and developing mutually beneficial human-wildland interactions that are compatible with the evolutionary history of those systems. Restoration is based on an ecosystem's reference conditions (or natural range of variability); the difference between reference conditions and contemporary conditions is used to assess the need for restorative treatments and to evaluate their success. Since ecosystems are highly complex and dynamic, it is not possible to describe comprehensively all possible attributes of reference conditions. Instead, ecosystem characteristics with essential roles in the evolutionary environment are chosen for detailed study. Key characteristics of structure, function, and disturbance-especially fire regimes in ponderosa pine ecosystems-are quantified as far as possible through dendroecological and paleoecological studies, historical evidence, and comparison to undisrupted sites. Ecological restoration treatments are designed to reverse recent, human-caused ecological degradation. Testing of restoration treatments at four sites in northern Arizona, USA, has shown promise, but the diverse context of management goals and constraints for Southwestern forest ecosystems means that appropriate applications of restoration techniques will probably differ in various settings.

The age structure in 1876, the last year of the natural frequent-fire regime, of an unharvested ponderosa pine forest in northern Arizona was reconstructed from living and dead dendrochronological samples. Approximately 20% of the trees were Ͼ200 yr old in 1876 with ages ranging to 540 yr. If dead trees had not been included in the reconstruction, the distribution would have been biased toward younger trees and a 40% shorter age range. The presettlement age distribution was multimodal with broad peaks of establishment, consistent with the model of regeneration in ''safe sites'' where herbaceous competition and fire thinning are reduced. Although fire disturbance regimes and climatic conditions varied over the centuries before 1876, a clear relationship between these variations and tree establishment was not observed. Due to fire exclusion, reduced grass competition, and favorable climatic events, high levels of regeneration in the 20th century raised forest density from 60 trees/ha in 1876 to Ͼ3000 trees/ha in 1992. An ecological restoration experiment initiated in 1993 conserved all living presettlement trees and reduced the density of young trees to near-presettlement levels. Two important components for evaluating the restoration treatment effects are monitoring of old-tree persistence and patterns of future regeneration in the context of the presettlement reference age structure.

No abstract