Skip to main content
Dryad

Data from: Carbon content of tree tissues: a synthesis

Data files

Abstract

Assessing the potential for forest carbon (C) capture and storage requires accurate assessments of C in live tree tissues. In the vast majority of local, regional, and global assessments, C content has been assumed to be 50% of tree biomass; however, recent studies indicate that this assumption is not accurate, with substantial variation in C content among tree species as well as among tissue types. Here we conduct a comprehensive literature review to present a global synthesis of C content in tissues of live trees. We found a total of 253 species-specific stem wood C content records owing to 31 studies, and an additional 34 records of species with C content values of other tissues in addition to stem wood. Stem wood C content varied significantly as a function of biome (tropical, subtropical/ Mediterranean, temperate/ boreal) and species type (conifer, angiosperm). Conifer species exhibited greater wood C content than angiosperm species (50.8 ± 0.7% (95% C.I.) vs. 47.7 ± 0.3%, respectively), a trend that was consistent among all biomes. Although studies have documented differences in C content among plant tissues, interspecific differences in stem wood appear to be of greater importance overall: among species, stem wood C content explained 37, 76, 48, 81, and 63% respectively of the variation in bark, branch, twig, coarse root, and fine root C content values, respectively. In each case, these intraspecific patterns approximated 1:1 linear relationships. Most published stem wood C content values (and all values for other tree tissues) are based on dried wood samples, and so neglect volatile C constituents that constitute on average 1.3 – 2.5% of total C in live wood. Capturing this volatile C fraction is an important methodological consideration for future studies. Our review, and associated data compilation, provides empirically supported wood C fractions that can be easily incorporated into forest C accounting, and may correct systematic errors of ~1.6 – 5.8% in forest C assessments.