Herbs and secondary woodiness – keeping up the cambial habit


The evolution of woodiness was a key developmental innovation among land plants. It permitted large changes in body-size, habit and functional complexity and influenced global carbon cycling. The ability of major clades to evolve variable growth forms and switch between herbs, shrubs, climbers and trees is of considerable ecological significance and in many groups involved significant changes in wood development. In this issue of New Phytologist, the finding by Lens et al. (pp. 12–17) that cambial development in Arabidopsis thaliana is increased by knocking out two genes implicated in flowering time provided a welcome insight for understanding how herbs retain the potential for woody growth and for understanding island woodiness – a phenomenon that has fascinated botanists for over a century.

‘… there is no point in preparing a well adjusted launch pad for seeds and their dissemination if there is no time or energy left to produce them!’

Woodiness has sometimes been a frustrating trait to study. For evolutionary biologists, different kinds of vascular cambium evolved in different major clades including unifacial cambia producing only secondary xylem, and bifacial cambia developing both secondary xylem and secondary phloem. Some plants, such as the fern Botrychium s.l., long believed to show evidence of wood has now been shown not to (Rothwell & Karrfalt, 2008). Furthermore, loss of wood coupled with high levels of morphological reduction has contributed to the difficulty of correctly establishing affinities of some groups such as the Hydatellaceae, that were previously classified among monocotyledons until their recent discovery as sister to the Nymphaeales (Saarela et al., 2007). For ecologists, attempts to study wood traits in relation to herbaceous habits can be exasperating (Eiten, 1991). Small-bodied, annual, biennial and even perennial herbs may produce a cambium, which may develop only tiny amounts of wood in only localized parts of the plant body. Herbs such as Arabidopsis thaliana have therefore retained the genetic and developmental potential to develop woodiness under the phenotypic guise of herbs. This kind of ‘hidden woodiness’ within lineages may be highly adaptive for species to survive and radiate during high levels of perturbation and periods of ecological upheaval – when times are bad – but also potentially preaptive for larger woody growth forms during more equable conditions when times are good (Rowe & Speck, 2005). The facility for adopting a herbaceous strategy has even been aired as one of the factors explaining how angiosperms replaced other seed plant lineages during their rise to dominance (Bond, 1989).

Comparisons of mutant organisms might not represent the exact changes in gene functioning that occur during selection and natural speciation events; however Lens et al. demonstrate how knocking out two MADS-box genes related to flowering – quantitatively a rather modest change of the genome – can produce a large anatomical change in cambial growth. Arabidopsis, like many other eudicots, is a woody herb as opposed to nonwoody herbs in which profound loss of cambial activity has occurred, such as in monocots, Nymphaeales and the diminutive Hydatellaceae. Profound loss lead to truly herbaceous organizations constrained to woodless life histories and morphological radiations without wood and secondary phloem. Such plants remained small, specialized and often aquatic, apart from some monocots that re-invented ways of producing large growth forms via massive, but frost sensitive, primary meristems in palms and novel kinds of secondary meristems in plants such as Yucca, Cordyline and Dracaena. The nonreturn to woody growth forms from truly herbaceous lineages indicates that the genetic and developmental changes had been irreversibly impaired or lost with no potential quick fix within the genome to develop wood and facilitate radiations of woody growth forms.

So what does the lingering presence of wood tissue in Arabidopsis mean for the life history of Arabidopsis? Does the small amount of wood produced serve any function or respond to changes in environmental conditions? A number of studies have described how different, sometimes rather unbiological manipulations can nudge wild-types or mutants into producing complete wood cylinders, but there have been few studies showing what this actually does for the plant in terms of stem hydraulics or mechanics – arguably the two main functions of wood. Wild-type A. thaliana produces small amounts of secondary xylem at the base of the floral axis, localized around fascicular parts of the stem but these do not form an entire ring. Reports on the biomechanics of the floral axis demonstrate that the larger volume of lignified, interfascicular tissue is the principal mechanical tissue (Jones et al., 2001). Mechanical perturbation, involving many repeated flexures per day, mimicking a windy environment, produce shorter, more flexible floral axes that mostly do not develop secondary tissue (Paul-Victor & Rowe, 2010). In contrast to this kind of mechanical stimulus, application of static loads (weights placed at the top of the stem) actually induce more secondary growth (Ko et al., 2004). These contrasting results suggest that the diminutive floral axis can nevertheless modify the degree of secondary growth with respect to different kinds of mechanical perturbation.

A fascinating aspect of the Lens et al. article is the suggestion that genes implicated in the timing of flowering are also responsible for blocking development of the main mechanical lateral meristem of the stem. Genetic integration of flowering and secondary growth development would seem to make biological sense. By the onset of flowering in a short-lived herb, it is probably better to mobilize resources for ensuring seed production rather than fine-tuning height and stem stiffness – there is no point in preparing a well adjusted launch pad for seeds and their dissemination if there is no time or energy left to produce them!

Timing plays a central role for development and evolutionary change in many organisms, but particularly in plants. The geometry and modular construction of vascular plants include apical and lateral meristems as well as primary and secondary tissues in linear branching structures. Younger distal bits of the plant are generally supported mechanically and hydraulically by older proximal bits so that changes in developmental timing can readily influence overall body size and form. Herbs, shrubs, climbers and trees can differ by several orders of magnitude of size, height and mass. Such big changes in form may be possible between putative ancestor and descendent, by relatively simple changes in developmental timing of one or more meristems and without the need for ‘inventing’ or ‘experimenting’ with new structures and new developmental pathways. This is at least partly why heterochrony is believed to be so influential in the evolution of size and form in plants. If lineages manage to hold on to the ability to develop secondary meristems even during radiations as herbs, they retain the key to radiating again, possibly rapidly, as woody plants.

While heterochronic shifts in plant organization have been popularly applied to wood development traits (Carlquist, 2009), other tissue complexes and combinations must be coupled with wood formation, from simply maintaining an outer ring of bark tissue around the wood cylinder to plumbing in hydraulic connections between the root, stem and leaves. Cambial growth can be mechanically complex: when wood cylinders expand within the primary body of a plant stem, self-harm will result unless self-repair mechanisms are available that keep pace with the expanding wood cylinder (Masselter & Speck, 2008). Interestingly, some of the earliest plants with wood from the fossil record show periderm development that actually resembles wound tissue rather than a compact organized, periderm layer (Rowe & Speck, 2004).

Comparisons across lineages of early angiosperms and eudicots have indicated the importance of radiations as herbs followed by secondary woodiness (Lens et al.). We suspect that suppression and re-expression of wood formation under diverse phylogenetic and environmental contexts has an even wider significance beyond cases of island or ‘island-like’ woodiness. Furthermore, the phenomenon probably varies on a case-by-case basis in terms of degree of reduction (complete loss or limited expression) as well as the specific traits involved (e.g. paedomorphic xylem and ray traits). Shifts towards herbs are particularly striking among early diverging angiosperm lineages, including groups in which wood expression disappeared altogether such as Nymphaeales, Hydatellaceae and monocots, as well as complex patterns of herbaceous and woody organization among magnoliid lineages, particularly the Piperales. Evolution of herbaceous growth forms and subsequent returns to woodiness occurred many times in angiosperms, probably under a wide range of ecological contexts. It is of great interest that genes implicated in flowering also regulate cambial growth (Melzer et al., 2008) implying that such a connection at the genetic level might be consistent with short, quick life histories in which flowering is forced within a narrow time frame.

So, if angiosperms can show such high levels of evolution towards herbs and then woodiness, why don’t gymnosperms? Apart from the small-bodied, but nevertheless woody earliest seed plants, only one report of a herbaceous gymnosperm exists from the fossil record and that is of a small bodied, fertile plant that nevertheless produced secondary xylem from a vascular cambium (Rothwell et al., 2000). Among extant gymnosperms, none develops a profoundly herbaceous growth form lacking a cambium altogether and even the famously parasitic Parasitaxus has a woody, shrub-like stem. The linking of SOC1 and FUL genes with the suppression of cambial activity in angiosperms might therefore be an important clue for explaining how angiosperms but not gymnosperms can switch to herbs and back to woody plants. As new comparative studies of regulatory genes in the major clades come to light (e.g. Becker & Theissen, 2003), their combination with parallel studies investigating size, form and function of seed plants might at last reveal why some clades can strategically suppress wood development or even get rid of it, while others have grimly held onto their woody bauplans, possibly curtailing their potential for competing as herbs. Furthermore, such approaches might also reveal the genetic, developmental and ecological links between flowering, woodiness, growth form and life history.