Aim: For trees, wood density is linked to competing energetic demands and therefore reflects responses to the environment. Climatic trends in wood density are recognized, yet their contribution to regional biogeographical patterns or impact on forest biomass stocks is not understood. This study has the following two objectives: (O1) to characterize wood density–climate trends for coarse (i.e., angiosperm versus gymnosperm) and fine (i.e., within-species) taxonomic units and test a predictive model that incorporates these trends into a model that assumes range-wide wood density is constant; and (O2) to assess the impact of climate-driven intraspecific variation on forest biomass stocks for major tree species. Location: We use an assemblage of eastern U.S. tree species for assessing climatic trends (O1), and then apply fitted models to forest inventory data spanning the eastern U.S.A. to assess impacts of forest carbon estimation procedures (O2). Methods: We compared hierarchical models fitted to the full data to characterize wood density/climate gradients and to assess the impact of within-species variation (O1). Then, we compared predictions of biomass stocks from the climate-variable model with those of the static model using the Forest Inventory and Analysis (FIA) database (O2). Results: We found among- and within-species trends related to temperature and moisture regimes, with differing responses between angiosperms and gymnosperms. Incorporating within-species variation in wood density increases the carbon stock of the study region by an estimated 242 Tg when compared with a species-only model. Main conclusions: Intraspecific variation in wood density across species ranges suggests that climate influences investment in stem wood within tree species and contributes to biogeographical patterns in wood density in the eastern U.S.A. This variation impacts forest biomass stock assessments, and thus contributes refinements to the U.S. National Greenhouse Gas Inventory. In addition, our work highlights the potential for combining trait data and forest inventory to infer forest ecological processes at broad spatial scales.
Bibliographical noteFunding Information:
We wish to thank Brad Oberle, Ethan Butler and Jake Grossman for their comments on this manuscript. This work is supported by the U.S. Forest Service Northern Research Station and the Minnesota Agricultural Experiment Station.
© 2017 John Wiley & Sons Ltd
- Bayesian hierarchical models
- biomass modelling
- climatic variation
- forest inventory
- stem wood density