External drag force (drag hereafter) exerted by wind or water flow is a major stress factor for plants in both aquatic and terrestrial environments, and has been implicated as a selective force determining the evolution of plant structural and morphological traits (Ennos, 1997). Plants can prevent mechanical damage from drag either through bending and reconfiguration thus minimizing the forces encountered (drag avoidance, hereafter) or by producing strong support structures that resist large forces (drag tolerance, hereafter) (Wainwright et al., 1976). The current view holds that these two strategies are negatively correlated; strong structures tend to be relatively rigid and thus less capable of reconfiguration. This fundamental trade-off imposes important constraints on the evolution of traits that determine both strategies for aquatic plants (Puijalon et al., 2011). In this issue of New Phytologist, the research of Butler et al. (pp. 137–149), however, challenges this view for terrestrial woody plants. They carried out a large-scale comparison on drag–trait associations amongst 39 woody plant species of tropical Australia and showed that these plants did not conform to the expected drag avoidance–tolerance trade-off. Their work thus sheds a new perspective on trait evolution in plant, terrestrial woody plants in particular.
‘New large scale comparisons on drag across plants differing in life-history will tell us how fluid flow constrains plant trait evolution and contributes to the divergence or convergence of the adopted growth strategies of plants.’
Quantifying drag forces on plants
Basic theory of fluids predicts that the drag (D, in N) on an object is given as:
with cd, the drag coefficient which inversely scales with the form driven streamline of the area A (in m2) perpendicular to the flow; ρ, the fluid density (in kg m−3); and U the flow speed of the fluid (in m s−1) (Vogel, 1994). Plants can reduce drag by bending and aligning their shoots and leaves, which reduces A and changes their streamlining expressed by cd. The contribution of this streamlining can be quantified by analyzing the relationship between log D and log U (D ∼ Uv ), with lower values of v indicating efficient streamlining and v =2 indicating a completely rigid object (Eqn 1) (Vogel, 1994).
Measurements of drag on terrestrial plants are typically done in wind tunnels. But most wind tunnels are insufficiently large to measure drag effectively on tree branches, let alone entire trees. This problem has been partially solved by using small-sized aero-elastic tree models that are dynamically realistic downscaled versions of trees (e.g. Gardiner et al., 1997), but this approach misses the link to differences in the suite of plant traits across species. Drag on aquatic plants due to water flow has been measured in so-called flumes: experimental gutters in which water flow can be controlled (Bouma et al., 2005). None of these approaches, however, are applicable to investigate the behavior of trees under wind drag in relation to plant traits across a variety of species. Butler et al. solved this problem by mounting 1.5-m tree shoots on a motor vehicle and by driving the vehicle at varying speeds producing air flow of 2.8 to 27.8 m s−1. The shoots were attached to a low friction stage connected to a load cell, and changes in the wind-exposed area (A) of shoots were monitored with a webcam mounted on the shaft of the vehicles front bumper. With this original and effective new methodology, Butler et al. demonstrated for the first time how plant traits contributed to drag resistance and avoidance across a wide range of woody plant species.
The strength vs streamlining trade-off revisited
A trade-off between drag avoidance and tolerance might be expected from mechanical constraints on the structure and tissue composition of plants. The ability to resist large forces (drag tolerance) is positively related to the thickness and break strength, indicated by the modulus of rupture (MR), of plant support structures (e.g. stems or petioles), though other traits such as root strength and rooting depth play a role as well (Gardiner et al., 1997). The thickness of a structure (expressed by the second moment of area) also entails a larger bending resistance (Niklas, 1992). In addition, in plants, MR tends to be positively correlated with tissue rigidity (expressed by Young’s modulus E ), especially in woody species where both tissue traits are closely correlated with wood density (e.g. van Gelder et al., 2006). Thus, thick, stiff and strong support structures tend to resist bending, and in that way impede reconfiguration and associated drag reduction (but see Vogel, 2009).
For aquatic plants, Puijalon et al. (2011) quantified drag tolerance (expressed as the maximum tensile force that plants can resist) and drag avoidance (expressed as 1/D with the D representing the actual drag experienced by plants in moving water) across 28 herbaceous species from submerged aquatic vegetation. They found the expected negative correlation between the strength to resist breaking by drag and the ability to avoid drag (Fig. 1a). Traits that were associated with drag avoidance were overall size, thin petioles and leaves, and low tissue rigidity, whereas large size, tissue strength and leaf and petiole thickness conferred drag tolerance. They argued that the existence of such a trade-off has important ecological and evolutionary consequences because it imposes constraints on the evolution of many morphological, anatomical and architectural traits. Selection for traits enhancing drag tolerance would go at the expense of drag avoidance, and vice versa.
Many studies implicitly assume that the drag tolerance–avoidance trade-off extends to terrestrial plants (e.g. Niklas, 1996; Anten et al., 2010) but this has not previously been tested (Vogel, 2009; Puijalon et al., 2011). Butler et al. found that woody terrestrial species with thicker more rigid stems exhibited a relatively smaller reduction in exposed area under wind drag. Unexpectedly, however, these species with rigid stems also exhibited more efficient streamlining: drag increased more slowly with increasing wind speed than for species with more flexible stems (Fig. 1b). Species with rigid stems were thus able to both resist larger forces and reduce drag more effectively by streamlining than species with more flexible stems. In other words, Butler et al. did not find support for the existence of a trade-off between drag tolerance and drag avoidance across terrestrial woody plants.
How can a larger bending resistance facilitate streamlining even if it impedes reconfiguration? Butler et al. proposed that trees with stiffer shoots tend to exhibit greater stability under wind loading and are thus less prone to so-called flagging. It has been shown that a flag-like polythene structure can have as much as a 10-fold larger drag coefficient than a completely rigid structure of the same shape (see Vogel, 2009), the magnitude of this effect depending on the geometry of the structure and the turbulence of the flow (Hoerner, 1965). The exact mechanisms responsible for this increase in drag are still under debate, but recent work suggests that a destructive amplification of vortices around a flagging structure plays a role (Argentina & Mahadevan, 2005). Flagging has also been implicated in relation to wind drag on individual leaves (Vogel, 2009). Butler et al. are the first to show that such mechanisms may also operate on branches of woody species, with potentially large implications for our view of selection on trait properties of trees worldwide.
Wind drag on trees vs hydrodynamic drag on aquatic plants
The question then arises as to why the results for aquatic plants in water flow and trees in wind were so different. Although aquatic and terrestrial plants differ in many aspects, the difference in the consequences of wind vs water flow may be a crucial one. Aquatic plants typically experience a 25-times larger drag than terrestrial plants exposed to a similar wind speed (Denny & Gaylord, 2002), but, of course, water is denser, more viscous and flows slower than air. The generally accepted way to determine the dynamic similarity between flow experiments under different conditions (such as water vs wind effects on plants) is to calculate the Reynolds numbers (Re), a dimensionless number that indicates the ratio of inertial to viscous forces. Flow tends to shift from laminar to turbulent with increasing Re values, while vortex shedding around objects in a flow also becomes more likely at high Re (Rott, 1990). We roughly calculated that the Re values in the water flow experiment of Puijalon et al. (2011) were 5–10-fold smaller (Re = 2 × 104 to 4 × 104) than during wind drag measurements of Butler et al. (Re c. 20 × 104). We suggest that with little flagging, water flow may have selected for aquatic plants with rigid structures that resist drag, and plants with flexible structures that avoid drag. By contrast, terrestrial plants face a greater impact of flow turbulence and the amplification of vortices by flagging. The study of Butler et al. shows that woody plants thus benefit from more rigid, sturdy branch structures to both tolerate drag and reduce drag, and may not show divergent drag tolerance vs drag avoidance strategies as do aquatic plants. Starting from such observations and the newly developed methodology by Butler et al., new large scale comparisons on drag across plants differing in life-history (terrestrial woody vs terrestrial herbaceous vs aquatic herbaceous) will tell us how fluid flow constrains plant trait evolution and contributes to the divergence or convergence of the adopted growth strategies of plants.