Rapid climatic changes have forced an intensive debate about the most effective carbon forestry approaches to mitigate human-induced greenhouse gas emissions. While related concepts such as the diversity-productivity relationship have been extensively investigated, effects of forest structural development on carbon dynamics remain less well understood. This is particularly true in the northeastern U.S., where secondary forests are recovering along multiple development pathways. The objectives of our study were (i) to identify structural development pathways in secondary northern hardwood-conifer forests, (ii) to analyze their differences in carbon storage, and (iii) to quantify the effect of structural drivers of carbon accumulation pathways.
Our focal ecoregion for this study was the Adirondack-New England mixed forest in the northeastern U.S. We conducted an intensive inventory of structural attributes at 45 mature, unmanaged forest sites. For these sites, we identified several distinct structural development pathways using Agglomerative Hierarchical Clustering (AHC). We employed a random forest algorithm to detect meaningful associations between carbon and structural attributes, and subsequently, to derive the partial effect of each structural variable on carbon storage.
Results/Conclusions
AHC suggested three different structural development pathways, including one softwood, and two hardwood-dominated clusters. Nine of the 19 variables investigated differed significantly between clusters. Among those were all variables related to the variability in tree dimensions and the heterogeneity of the canopy. The structural development pathways differed significantly in carbon storage with the highest amount of carbon stored in the softwood cluster. Nine variables were identified to be meaningful for carbon storage. Of those seven had a positive, one a negative, and one a bimodal relationship with carbon. The positive effect of structure was more distinct in hardwood compared to softwood-dominated forests. Live basal area dominated the effect of structural variables on carbon, followed by the heterogeneity of tree diameters, and the percentage of conifers.
Our study reveals the importance of accounting for forest structure in management approaches aiming to improve carbon storage to mitigate climate change. The variability of carbon stored in hardwood-dominated clusters indicates a particularly high potential to increase carbon in these forests. Forest management systems designed to enhance stand structural complexity are likely most effective to optimize carbon storage in production forests.