Stand age, disturbance history and the temporal stability of forest production.

• Upper Great Lakes forests exhibit high overall production temporal stability. • Production temporal stability varied by >2-fold among stands. • More severely disturbed forests displayed greater temporal production stability. • Stand structure and leaf traits were inconsistent predictors of temporal stability. • Properties conferring greater production rates and temporal stability may differ. Sustaining the terrestrial carbon (C) sink requires knowledge of the forest properties supporting stable production under increasingly variable climate conditions. We examined how stand disturbance history and age, structural complexity and species diversity, and leaf properties relate to the 10-yr stability of above-ground wood net primary production (NPPw) in northern temperate forests of Michigan, USA. Our investigation centered on separate deciduous, evergreen, and mixed late successional stands initiated over a century ago and free of recent disturbance, a "Cut Only" chronosequence established following clearcut harvesting, and a "Cut and Burn" chronosequence that regenerated following experimental clearcut harvesting and fire. The temporal stability of stand production was calculated from the 10-yr coefficient of variation (CV) of annual NPPw estimated from tree cores; canopy rugosity, a measure of structural complexity, was estimated using terrestrial LiDAR; and >1500 subcanopy leaves were sampled for leaf mass area and chlorophyll fluorescence intensity. The temporal stability of stands differed by >2-fold, from 5% to 11% CV of NPPw. Counter to expectations, we found that NPPw stability was greatest in the more severely disturbed Cut and Burn stands and lowest in late successional stands. Despite similar successional patterns of species diversity and structural complexity, NPPw stability increased in Cut Only stands and declined in Cut and Burn stands as age, diversity and canopy rugosity increased. The NPPw of more diverse, late successional deciduous forests was more temporally stable than that of evergreen forests. We conclude that management for maximal rates of production may not confer temporal stability, indicating future studies are needed to elucidate the stand and canopy properties that support both high production rates and stability. [ABSTRACT FROM AUTHOR]