Wood density (WD) is a complex trait that integrates wood anatomical, structural, and physiological properties.  Trees with higher wood density tend to be less vulnerable to xylem cavitation and implosion and more resistant to buckling stresses, enabling persistence in water-limited or exposed habitats.  Given its relationship with plant function, one may expect WD to be evolutionarily conserved, but WD also exhibits environmental plasticity.  However, it is not clear how WD varies within species due to environmental influences or across species due to evolutionary history.  Our objectives are to (1) obtain rigorous estimates of WD, normalized for environmental factors, for 305 US tree species, and (2) evaluate environmental versus evolutionary influences on species-specific WD values.  WD values and important covariates (e.g., latewood proportion, DBH, tree age, wood type) were extracted from the literature and climate data were obtained from the CRU database. We implemented a Bayesian meta-analysis to synthesize 1741 WD values, which allowed us to simultaneously incorporate the effects of climate and covariates and to estimate missing/unreported covariate data.  Within the model, we specified a hierarchical model for species-specific, normalized WD values based on a phylogeny for the 305 species, which facilitated estimation of WD for data-poor taxa.

Results/Conclusions

Wood type had a strong effect on species-specific WD.  For example, softwood species had the lowest WD (mean = 0.46 g/cm³; range = 0.30 (Thuja plicata) to 0.58 (Juniperus silicicola)); oaks, which support both ring- and diffuse-porous wood, had the highest WD (mean = 0.56 g/cm³; range = 0.11 (Quercus laevis) to 0.82 (Q. wislizeni)); ring-porous species had the second highest, and least variable WD (mean = 0.54 g/m³; range: 0.27 (Sapindus drummondii) to 0.70 (Magnolia acuminata)); and, diffuse-porous species ranked third (mean = 0.49 g/m³; range = 0.11 (Olneya tesota) to 0.74 (Fagus grandifolia)). Q. wislizenii had the highest WD value, which was 7.6 times higher than the lowest value of 0.11 g/cm³ for Olneya tesota and Q. laevis. WD clearly differs between species partly due to wood type, but it was also affected by exogenous factors such as precipitation and endogenous, tree-level factors such as growth rate. Within a given species, WD is lower in high-precipitation sites and in fast growing trees, indicating some degree of environmental or phenotypic plasticity. Once normalized for environmental and endogenous influences, species-specific WD values were significantly different between many of the species, thereby also demonstrating the role of evolutionary history.