Click below for larger images and slideshow more...

A tangential section of the wood of the lobelioid Cyanea coriacea.

A radial section of the wood of Cyanea coriacea confirms that all ray cells are upright.

The “rosette tree” lobelioid Delissea undulata in it native habitat in Hawaii. This species is now extinct.

A tangential section of wood of Delissea undulata shows scalariform pitting in a lateral vessel wall. Such pitting may represent an extension of the scalariform pitting on vessels in metaxylem into the secondary xylem.

Graph showing how vessel element lengths change over time for a typically woody species, Eriobotrya japonica, and for two “woody herb” species, Talinum guadalupensis (now a species of Cistanche) and Macropiper excelsum.

Ray type evolutionary diagram, in which the paedomorphic types are added to the types found in typically woody plants. There is no sharp distinction among the types, which are, however, exceptionally valuable to understanding how wood evolves.

PAEDOMORPHOSIS

 PAEDOMORPHOSIS [ PDF ]

   My 1962 paper on paedomorphosis in woods is a good example of concern with general evolutionary phenomena.  What I would like you to remember in looking at that essay is that in the first half of the 20th century, botanists believed that angiosperms started as woody plants in the tropics and that less woody lines, culminating in annuals, developed as these phylads progressed into temperate areas.  The woodier, the more primitive—that was the accepted wisdom.  Nobody seemed willing to say that woodiness could be secondary, or that woodiness could increase or decrease readily during the evolution of any particular group of plants.  Amazing!  But I saw in Asteraceae how that enormous family has been very plastic in terms of form.  In southern California, genera such as Haplopappus, Ericameria, Chrysothamnus, Senecio, and many others show secondary woodiness.  Frost is relatively minimal in southern California, so why should a plant die back to the ground or even die back to seeds each year?  That’s not economical if the plant can survive a few degrees of frost.  Woody mints (Lamiaceae) and buckwheats (Polygonaceae) also show this in southern California.  And, of course, secondary woodiness occurs on islands (see Secondary Woodiness under the Island Biology heading). 

   The woods of these plants (and of annuals themselves) showed prolonged juvenilism: more upright ray cells than expected, longer vessel elements.  The length of vessel elements was at that time considered a good measuring stick of primitiveness: the longer the vessel element, the more primitive.  But I saw how the cambia in herbs and secondarily woody plants were working.  Instead of rapidly subdividing to form shorter cells at the onset of secondary growth, as in truly woody species, these “woody herbs” were keeping the juvenile patterns indefinitely: for the life of the plant in many cases.  The juvenile ray types didn’t even fit into the Kribs scheme for ray types (the sampling of Kribs had been based on truly woody plants, of course).  Groups like the Hawaiian lobelioids (e.g., Cyanea) had rays consisting only of upright cells for the length of life of the plant, or maybe eventually a few procumbent cells, but not many.  Rays like this didn’t correspond to descriptions of rays in literature on wood anatomy.  The heavy influence of forestry schools had skewed the world’s view of wood anatomy.  Nobody had studied wood of plants that people couldn’t use to build furniture or houses from!  Why not?  Weren’t all woods worth studying?  There are many, many woody plants that can’t be used for building houses or furniture.  Certainly from an evolutionary point of view all woods are worth studying.  But forestry schools had a mission to know about useful plants, and unwittingly, forestry schools have dictated that useless plants should not be studied with respect to wood anatomy.  There was little interest in wood anatomy outside of forestry schools during my early years as a graduate student.  However, I decided I could study wood anatomy—couldn’t anyone do that regardless of whether they were affiliated with a forestry school?   The Kribs ray types, useful to major compendia like Metcalfe and Chalk’s (1950) “Anatomy of the Dicotyledons” had to be expanded, and so I designated paedomorphic ray types to show modification, by juvenilism, of the main Kribs ray types.  Also, pitting types in vessels, as in the tree Begonia parviflora and in the Hawaiian lobelioids, stayed juvenile (scalariform lateral wall pitting) instead of changing to opposite or alternate circular bordered pits as happens during secondary growth in typical woody plants. 

   When I graphically compared vessel element length to accumulation of secondary xylem, I found that “woody plants” such as Eriobotrya and “woody herbs” such as Macropiper and Talinum (now Cistanche) had different curves: protracted juvenilism in the woody herbs. I happened to choose those because I could measure vessel element lengths and see ray cell shape easily in radial sections of them.  Instead of rapid decrease in vertical length of ray initials and of fusiform cambial initials as in the woody plants (a dip followed by increase in fusiform cambial initial length), the ray initials and fusiform cambial initials of herbs and “woody herbs” divided transversely very slowly over time, sometimes hardly at all, and there was no increase in length after the dip in the curve.  Paedomorphosis was the correct word to choose for this, and I chose it.  The word paedomorphosis was invented to refer to an individual that reaches adulthood (sexual maturity) still having juvenile some characteristics.  A plant that can bear flowers and seeds while retaining characteristics it had as a seedling can thus be described with this term.  As with many new ideas, my concepts on paedomorphosis in wood were opposed.  They were widely misinterpreted, also—it’s easy to mistrust an idea if one doesn’t understand it. To be sure, DNA studies eventually supported my ideas, and in fact, all workers in wood anatomy now have accepted my ideas of paedomorphosis in wood.

   Paedomorphosis occurs in numerous woods, mostly those derived from nonwoody or less woody ancestors.  However, it can occur in wood of herbs themselves (yes, most herbs can and do have some wood—the difference between a woody dicotyledon and an herbaceous dicotyledon is only one of degree.  And not surprisingly, there are different degrees of protracted juvenilism in different woods.  Recognition that there is heterochrony in woods makes understanding them more difficult, it’s true.  However, it’s better than pretending that all woods have the same timing in their departures from juvenilistic patterns.  Degrees of wood juvenilism have not yet received recognition in manuals or glossaries used for teaching about wood anatomy.  Juvenilism or heterochrony in wood structure represents a fund of traits that can be accessed singly or together as a source for evolutionary change.  Thus, anyone interested in wood evolution ought to be fascinated by paedomorphosis.

   The obvious question posed by the phenomenon of paedomorphosis in wood is why it should exist in particular plants.  Or—perhaps more importantly—what the function of rapid onset of adult wood patterns is in the typically woody species.  Certainly paedomorphosis represents a dimension available to an evolving group of plants, an easy mechanism for pattern shift that probably requires simple genetic changes rather than long-term programs of change in cell size and shape. 

   Long fusiform cambial initials produce longer libriform fibers (as well as longer vessel elements).  Vertically longer ray initials produce upright ray cells.  Upright ray cells are related to vertical conduction rather than horizontal conduction.  (What no botany text mentions: the direction of elongation of a conductive cell is its direction of conduction, because elongation is a way of producing fewer cross-walls that would impede conduction. A series of square cells would provide more impedance than a single elongate cell of the same length).  Plants with wood paedomorphosis are limited in stem diameter, so radial conduction of photosynthates in rays is not as pronounced a function as in woodier plants.  Woods with paedomorphosis tend to grow in moister places or be succulents, so the value of long libriform fibers is operative in Eriobotrya japonica, but not in Talinum or Macropiper.  Scalariform lateral wall pitting of vessels (rather than circular bordered pits) is not an optimally strong design, and its occurrence in succulent stems is not unexpected.  In these examples, paedomorphosis represents a kind of release from mechanical strength considerations. There is no single pattern that applies to all woods with heterochrony, but one should go beyond mere description and examine the evolutionary significance of these patterns.  A curious case is formed by small shrubs that have uniseriate rays only, rays composed of upright cells only—the Paedomorphic Type III here.  I have described this type in my papers on Empetraceae [ PDF ], Myrothamnaceae, Tremandraceae, and the genus Empleuridium.  Such rays may also be found in some Epacridaceae.

   In its simplest form, paedomorphosis in wood can be said to be a continuation of patterns characteristic of herbs, because the “woody herbs” are not markedly different from herbs in their adaptational characteristics.  A rapid change from juvenile to adult patterns of wood structure should be taken as an indication that juvenile patterns are adaptive for a short period of time in the stem of a typically woody plant.  In turn, the interest here is what the difference is and why it takes place.

   The 1962 paper, predictably, was only a beginning in understanding how the phenomenon of paedomorphosis in wood provides a fund of outcomes that has character4ize so much of the woodiness we see today. The 2009 paper on xylem heterochrony showed that there is a range of paedomorphic expressions, ranging from accelerated adulthood (typical tree) to permanent juvenilism (monocots) in xylem development. If early angiosperms were indeed minimally woody as has been widely inferred in recent times, then acceleration of adulthood in xylem structure was necessary to cross over into woodiness. This can happen innumerable times in various clades, as can the reverse, production of non-woody (herbaceous) growth forms. Increase or decrease in juvenilism thus provides the explanation for the numerous growth forms we see today. The triggering of less or more juvenilism in xylem production is a uniquely angiospermous feature, and may be one of the main reasons for their success. Conifers have virtually no juvenilism in xylem ontogeny.

    The 2012 paper, "More Woodiness/Less Woodiness" extends the reach of this topic further, showing stages in progress from woodier to less woody or the reverse. Can we detect the direction? Within a given species, are degrees of woodiness actively evolving? Within a genus? What lines of evidence do we use? These and many more questions are considered.