Clearcutting a forest ecosystem can result in a drastic reduction of stand productivity. Despite the severity of this disturbance type, past studies have found that the productivity of young regenerating stands can quickly rebound, approaching that of mature undisturbed stands within a few years. One of the obvious reasons is increased leaf area (LA) with each year of recovery. However, a less obvious reason may be the variability in species composition and distribution during the natural regeneration process. The purpose of this study was to investigate to what extent the increase in gross ecosystem productivity (GEP), observed during the first 4 years of recovery in a naturally regenerating clearcut stand, was due to (i) an overall expansion of leaf area and (ii) an increase in the canopy's photosynthetic capacity stemming from either species compositional shifts or drift in physiological traits within species. We found that the multi-year rise in GEP following harvest was clearly attributed to the expansion of LA rather than a change in vegetation composition. Sizeable changes in the relative abundance of species were masked by remarkably similar leaf physiological attributes for a range of vegetation types present in this early-successional environment. Comparison of upscaled leaf-chamber estimates with eddy-covariance-based estimates of light-response curves revealed a broad consistency in both maximum photosynthetic capacity and quantum yield efficiency. The approaches presented here illustrate how chamber- and ecosystem-scale measurements of gas exchange can be blended with species-level LA data to draw conclusive inferences about changes in ecosystem processes over time in a highly dynamic environment.