Still rethinking the value of high wood density

? Premise of the study: In a previous paper, we questioned the traditional interpretation of the advantages and disadvantages of high wood density (Functional Ecology 24: 701-705). Niklas and Spatz (American Journal of Botany 97: 1587-1594) challenged the biomechanical relevance of studying properties of dry wood, including dry wood density, and stated that we erred in our claims regarding scaling. ? Methods: We first present the full derivation of our previous claims regarding scaling. We then examine how the fresh modulus of rupture and the elastic modulus scale with dry wood density and compare these scaling relationships with those for dry mechanical properties, using almost exactly the same data set analyzed by Niklas and Spatz. ? Key results: The derivation shows that given our assumptions that the modulus of rupture and elastic modulus are both proportional to wood density, the resistance to bending is inversely proportional to wood density and strength is inversely proportional with the square root of wood density, exactly as we previously claimed. The analyses show that the elastic modulus of fresh wood scales proportionally with wood density (exponent 1.05, 95% CI 0.90-1.11) but that the modulus of rupture of fresh wood does not, scaling instead with the 1.25 power of wood density (CI 1.18-1.31). ? Conclusions: The deviation from proportional scaling for modulus of rupture is so small that our central conclusion remains correct: for a given construction cost, trees with lower wood density have higher strength and higher resistance to bending.