Wood density under a microscope. By Kasia Ziemińska

Wood density—dry mass per fresh volume (g cm-3)—is one of the most commonly assesed trait in plant ecology and ecophysiology. It is easy to measure and it indicates the amount of carbon a plant invests in wood, which is an important component of plant carbon budget. At the same time, wood density is just one number, behind which hides a complex anatomical structure. And the relationship between wood density and its underlying anatomical components is not so straightforward.

So, what is inside wood? Angiosperm wood is composed of three main tissues: fibres, parenchyma and vessels (Fig. 1; there are some other ones, too, which occur less frequently or in low proportions, for example, tracheids). Gymnosperm wood has very different structure and is not considered here.

Figure 1. Cross-section through twig wood of Cornus kousa (Korean dogwood). Photo: K Ziemińska.

Because the density of cell wall material is approximately constant across species (~1.5 g cm-3 dry mass per dry volume 1), we can expect that wood density depends on the fraction of wall vs lumen. And indeed it does 2–4. But which particular tissue drives density variation across species: vessels, parenchyma or fibres?

Generally, wood density tends to be higher in species growing in dry environments as oppose to wet ones (there are exceptions, of course) 5. Hence, researchers hypothesized that high wood density would result from small vessel lumen fraction (low water availability, therefore low fraction of water conducting vessels). However, it turns out that wood density does not depend on vessel lumen fraction, as shown in a global analysis of >500 species 6.

Among the three primary wood tissues, fibre properties are the main drivers of wood density variation 2,4,7 but, really, the relationship between wood density and anatomy is triangular 3 (Fig. 2). One dimension of variation stretches along the fibre wall and lumen fractions, where high density species have high fibre wall and low fibre lumen fractions. The other dimension of variation is largely independent of wood density meaning that the same wood density can be achieved through diverse anatomies. This spectrum stretches along the fibre-parenchyma trade-off. That is, species with high parenchyma fraction and low fibre fraction are on one end of the spectrum and on the other end—species with low parenchyma fraction and high fibre fraction (but with large fibre lumen, Fig. 2b). Moreover, the lower wood density, the wider spectrum of anatomical options; while the higher wood density the less diverse anatomies converge towards high fibre fraction composed of thick-walled fibres. To use an analogy, take human adult body weight. Adults with lower weight can be tall and skinny or short and overweight. But people with very large weight will always look overweight.

Figure 2. A scatterplot (A) and a diagram (B) illustrating triangular relationship between wood density and anatomy. “(A) Black symbols represents 69 species from this study and grey symbols represent 24 species from Ziemińska et al 2013. Symbol diameter is proportional to species wood density. With the smallest diameter corresponding to the lowest density and the largest diameter to the highest density. Isolines indicate total fibre fraction increasing from left to right by a step of 0.1. Grey numbers above the X axis correspond to total fibre fraction indicated by a given isoline. (B) The six squares symbolize cross-sections through six different woods. Vessel fraction was relatively small, and did not show large variation across species, so for simplicity was omitted in this diagram. Wood density increases towards the top of the diagram. Total fibre fraction (brown wall + yellow lumen) and total parenchyma fraction (green, includes axial + ray) covary negatively with each other and approximately orthogonally to wood density.” From Ziemińska et al. 2015.

The finding that variation largely orthogonal to wood density is stretched on fibre-parenchyma spectrum is important and insightful but it should be noted that it was based on tropical species. It capitalized on the fact that vessel lumen fraction in tropical species sits on the lower end of the spectrum relative to the global variation and its range is fairly small (Fig. 3) 8. But temperate species encompass entire global variation in vessel lumen fraction, which is approx. double that of tropical species. So new framework describing anatomical variation independent of wood density needs to be developed that would include also temperate species (I’m working on it).

Figure 3. “Ternary plot showing the contribution of the three principal tissues in wood: the total amount of ray and axial parenchyma (RAP, %), vessels (V, %), and fibres (including tracheids) (F, %). Each dot represents a specimen. The trade-off between RAP and F tissue fractions shows a strongly negative correlation for all data and is especially clear in the tropical biome (see Table 2). Specimens are grouped according to three major climatic zones, which follow Köppen (1936).” From Morris et al. 2016.

So does understanding of anatomical drivers of wood density help us to understand biological meaning of wood density? Well, it seems that current knowledge does not give definitive answers but it generates new interesting research directions. For example, given that many hydraulic traits correlate with wood density (e.g., capacitance and midday water potential) and fibres mainly drive wood density, new questions arise: do fibres contribute to hydraulic functions or is the relationship between hydraulic traits and wood density correlative only, caused by a third, unknown factor? Fibre-parenchyma trade-off is particularly interesting because fibres are usually dead and parenchyma is alive. Hence, per given wood density, some species have lots of respiring tissue able to store and transport carbohydrates, while other species devote that space to dead fibres. What could be the costs and benefits of one strategy versus the other is an intriguing question and remains unanswered.

Wood density is a direct outcome of wall and lumen fractionswhile lumen diameter and wall thickness will only have an indirect effect modified by the number of cells (or the volume they occupy in wood, to be exact). For example, low wood density can result from both: few, thick-walled fibres or many, thin-walled fibres (Fig. 2). The same logic applies to vessel lumen (or any cell lumen, really). This is why the relationships between wood density and vessel lumen diameter are not causative and require caution when interpreting them. This is also the reason why studies reporting negative correlation (e.g., 9) or no correlation (e.g., 4) are not necessarily contradictory—both outcomes are perfectly plausible. It is then interesting to ask: what is driving presence or absence of correlations in these various studies? I suspect the more diverse species group and/or the smaller environmental gradient studied, the more likely no correlation would be observed. But this hypotheses needs to be tested. Another relevant here trait is wall-to-lumen ratio, also called “implosion resistance” (or “wall thickness-to-span ratio”7 or “thickness-to-diameter ratio”10), calculated as the thickness of walls of two neighbouring conduits (t) to conduit diameter (b) squared (t/b)2 first proposed by 11. That study suggested a mechanistic link between P50 (water potential at which 50% of conductivity is lost) and wood density through the wall-to-lumen ratio parameter. And this idea has penetrated literature and keeps on being repeated (e.g., 12). However, variation in wall-to-lumen ratio does not cause variation in wood density, at least not in angiosperms (it is somewhat different in gymnosperms) for the reasons I mentioned above: cell wall thickness or lumen diameter are not sole traits driving wood density.

So what does it all mean? Altogether, these findings strongly imply that the same wood density, does not necessarily correspond to a single functional strategy, and, especially in lower wood density species, there is a wealth of strategies independent of carbon investment in wood. This line of research also shows that anatomy offers an indispensable insight into the ‘black-box’ functional traits such as, in this case, wood density.

Variation in anatomical structure is enormous and traits described here are only the tip of an iceberg. I did not talk about cell sizes (except for vessel diameter), cell shapes, pits, tissue connectivity, pith, phloem or bark…and the list could go on and on. The functional and ecological meaning of much of anatomical variation is an unchartered and exciting territory for future research.

Kasia Ziemińska
23 October 2019


  1. Kellogg, R. & Wangaard, F. Variation in the cell-wall density of wood. Wood Fiber Sci. 1, 180–204 (1969).
  2. Ziemińska, K., Butler, D. W., Gleason, S. M., Wright, I. J. & Westoby, M. Fibre wall and lumen fractions drive wood density variation across 24 Australian angiosperms. AoB PLANTS 5, (2013).
  3. Ziemińska, K., Westoby, M. & Wright, I. J. Broad Anatomical Variation within a Narrow Wood Density Range—A Study of Twig Wood across 69 Australian Angiosperms. PLoS ONE 10, e0124892 (2015).
  4. Fortunel, C., Ruelle, J., Beauchêne, J., Fine, P. V. A. & Baraloto, C. Wood specific gravity and anatomy of branches and roots in 113 Amazonian rainforest tree species across environmental gradients. New Phytol. 202, 79–94 (2014).
  5. Chave, J. et al. Towards a worldwide wood economics spectrum. Ecol. Lett. 12, 351–366 (2009).
  6. Zanne, A. E. et al. Angiosperm wood structure: global patterns in vessel anatomy and their relation to wood density and potential conductivity. Am. J. Bot. 97, 207–215 (2010).
  7. Martínez-Cabrera, H. I., Jones, C. S., Espino, S. & Schenk, H. J. Wood anatomy and wood density in shrubs: responses to varying aridity along transcontinental transects. Am. J. Bot. 96, 1388–1398 (2009).
  8. Morris, H. et al. A global analysis of parenchyma tissue fractions in secondary xylem of seed plants. New Phytol. 209, 1553–1565 (2016).
  9. Pfautsch, S. et al. Vessel diameter and related hydraulic traits of 31 Eucalyptus species arrayed along a gradient of water availability. Ecology 97, 1626–1626 (2016).
  10. Jacobsen, A. L., Ewers, F. W., Pratt, R. B., Paddock, W. A. & Davis, S. D. Do xylem fibers affect vessel cavitation resistance? Plant Physiol. 139, 546–556 (2005).
  11. Hacke, U. G., Sperry, J. S., Pockman, W. T., Davis, S. D. & McCulloh, K. A. Trends in wood density and structure are linked to prevention of xylem implosion by negative pressure. Oecologia 126, 457–461 (2001).
  12. McCulloh, K. A., Domec, J.-C., Johnson, D. M., Smith, D. D. & Meinzer, F. C. A dynamic yet vulnerable pipeline: Integration and coordination of hydraulic traits across whole plants. Plant Cell Environ. 42, 2789–2807 (2019).

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s