Wood carbon fractions influence our understanding of forest carbon stocks at national scales.

Characterizing forest carbon storage and fluxes accurately is key for developing precise knowledge about the role forests play in mitigating climate change. However, many of the world’s most important studies and models likely overestimate forest carbon stocks, because they commonly overlook variability in tree species wood chemistry. Using one of the world’s largest forest inventory databases (i.e., the National Forest Inventory of the United States), alongside open access wood chemistry datasets, we show that accounting for variability in wood carbon concentrations in forest carbon models, significantly influences our understanding of forest carbon stocks from individual tree- to continental-scales. Models with species-specific wood carbon concentrations downward-revise estimates of forest carbon storage in the continental United States by 3.55% or ~ 0.65 Pg of carbon: a value equivalent to the entire forest carbon stock in the interior of Alaska. Accounting for species differences in wood chemistry, especially in angiosperms, would therefore suggest forest carbon stocks in the United States, and forests globally, are lower than previously estimated.