Click below for larger images and slideshow more...

Welwitschia mirabilis in its natural habitat near Swakopmund, Namibia. As is well known, Welwitschia plants have only two leaves (after their cotyledons); those probably correspond to left and right sides of the picture. The leaves split into fragments as the plant grows. Leaves grow at their bases.

Fenced in to protect it is the oldest known Welwitschia plant, dated to be 1300 years old.

A Welwitschia plant in which the two leaves have split relatively little. Inflorescences bearing old female cones can be seen attached to the plant where the leaves join the corky drumlike stem.

A female cone of Welwitschia reaching maturation. A few of the tubelike inner integuments, responsible for pollination of the ovules, protrude from the cone.

Shattered female cones (strobili) of Welwitchia. Papery bracts are attached to the seeds, which are about the size of pine seeds. The papery bracts may carry seeds along the Namib Desert in gusts of wind.

A male cone of Welwitschia. One of the units is at pollen-shedding stage: a ring of about six “stamens” (microsporophylls) surrounds the integument of the sterile ovule. The word strobilus is preferred to cone by many

A transection of a root of Welwitschia. Brown strands of phloem fibers are seen in the tan-colored background. The phloem fibers bulk relatively large in comparison to wood increments, which are not visible at this magnification.

A transection of the wood of Welwitschia. The fascicular areas consist of vessels intermixed among tracheids. These fasicular areas are separated from each other by rays. A couple of fibrosclereids, which stain dark purple, are seen at right.

A tangential section of the wood of Welwitchia The tracheids and vessels of fascicular areas stain dark green in this preparation. The rays have thin greenish walls; the rays separate the fascicular areas.

An SEM photograph of part of a Welwitschia fibrosclereid, showing the calcium oxalate crystals embedded in the surface.

A plant of Ephedra pedunculata growing, like a woody vine, up a telephone pole in Del Rio, Texas.

Eohedra torreyana is the most common shrub in Monument Valley, on the Utah-Arizona border.

Ephedra trifurca, up to 3 m high, occurs on the sand dunes of Imperial County, California.

Male cones (left) with stamens and female cones (strobili) of Ephedra viridis. The tubelike inner integuments of the ovules, now becoming seeds, are visible.

An enlargement of a male strobilus of Ephedra viridis shows that the microsporophylls form a kind of tube with the anthers radiating from its tip.

Papery bracts surround a seed of Ephedra trifurca. As they unfold (left) they form bladelike structures that catch the wind, carrying across the desert surface.

As the female strobilus of Ephedra andina (Cuzco, Peru) matures, the bracts become fleshy and red (probably attracting frugivorous birds) and the seeds become black.

A transection of the wood of Ephedra pedunculata shows growth rings. Vessels are a little smaller in the latewood than in the earlywood of each growth ring, but they are about equally abundant throughout.

A transection of the wood of Ephedra coryi shows strongly marked growth rings. Vessels are lacking in the latewood, but present in the earlywood. Tracheids are conductively safer than vessel elements, so an all-tracheid latewood prevents spread of air embolisms from one conducting cell to another, and preserves the water columns even if vessels in the earlywood embolize.

A transection of wood of Ephedra rupestris, a high-Andean species. The wood is almost devoid of vessels, a few of which are present at the bottom of the photo.

Pits in an Ephedra tracheid: the central torus, connected by strands with the edges of the pit cavity, is a feature Gnetales share with woods of conifers.

The perforation plate of an Ephedra torreyana vessel. The perforations are circular and bordered, quite different from the scalariform pattern basic to the pattern in flowering plants.

The perforation plate of Ephedra kokanica consists of perforations relatively large for the genus.

A longisection of a vessel of Ephedra torreyana. Helical thickenings on vessel walls are conspicuous in most New World species of Ephedra, but are rare in the Old World species. This feature was not reported in earlier studies on Ephedra wood. It is a prominent feature of numerous dicot woods in dry or cold areas.

Crystals fill the spaces between tracheids of Ephedra trifurca. These crystals are probably a mechanism to prevent beetles from eating the wood and bark of Ephedra.

Crystals line the spaces among ray cells in this section of Ephedra ochreata. The crystals are tiny, and can be seen vaguely with a light microscope, but a scanning electron microscope renders their form precisely.

Two trees of Gnetum gnemon on the University of Malaya campus, Kuala Lumpur.

The inflorescence of Gnetum gnemon is branched. Leaves of the tree are opposite and broad in shape. Only two species of Gnetum are trees—the other Gnetum species are lianas.

A closeup of an inflorescence if Gnetum gnemon shows the male “flowers” which have one microsporophyll or two fused ones. The ovules are sterile.

A female inflorescence of Gnetum gnemon. The ovules are clustered at nodes on the branches.

One of the ovules on this inflorescence of Gnetum gnemon has matured; the fleshy covering has turned red-orange.

Gnetum latifolium is a liana species. This inflorescence has numerous pointed seeds, close to maturation in this picture,

A transection of the wood of Gnetum gnemon shows vessels scattered through a background of tracheids.

Gnetum schwackeanum is a liana with successive cambia. Portion of two increments are shown here. Each cambium produces secondary phloem (which contains numerous fibers, bluish in color here) and secondary xylem (purplish) with vessels scattered in a background of tracheids.

The wood of Gnetum gnemon mostly has vessels with large simple perforation plates, but smaller vessels, like this one, viewed with SEM, can have perforation plates composed of numerous perforations like those of Ephedra.

A vessel from a maceration of wood of Gnetum gnemon shows four perforations, indicating a simplification compared with the numerous perforations seen in narrow vessels.

The pits on lateral walls of Gnetum are vestured (pits of G. gnemon shown here). This feature has arisen in Gnetum separately from its several origins in angiosperms.

Tori occur on pits of tracheids and vessels of some Gnetum species (Gnetum leyboldii shown here). Although the tori are not so pronounced as those of Ephedra, they clearly are present. So this feature in Gnetum must be added to others that demonstrate affinity with conifers, rather than flowering plants.

The helical thickenings in primary xylem of Gnetum gnemon (and other Gnetales) have circular bordered pits in the helices. This is a conifer feature not reported in angiosperms.



   The relationships of Gnetales have long been controversial, but few botanists are familiar with them on the basis of field work.  Because Gnetales are a not a major group in the minds of most botanists, and are not a group conspicuous in the world’s flora, they tend to be forgotten.  Actually, if the origin of angiosperms is an “abominable mystery,” so is that of Gnetales.  One would have thought that wood anatomy of such a curious group as Gnetales would be something long ago familiar, but I soon discovered that what we knew was based on examination of a few species.  My nine papers on Gnetales therefore do break new ground.  Before my work, conclusions on wood anatomy of Gnetales were based only on a few species, so the conclusions read by everyone had a narrow informational base.  I assembled wood of a surprisingly large number of species of the order.  I did an around-the-world trip to collect material of Gnetum in Malaya and Welwitschia in Namibia in 1989.  The visit to Malaya was funded by the National Geographic Society and the field work on Welwitschia was possible because of funds from the American Philosophical Society.  And when the national biological meetings were in Fort Collins, Colorado, I drove there and back from Claremont, collecting Ephedra woods throughout the western US (and seeing alpine wildflowers in the Rocky Mountains).  I even did a special air trip to San Antonio, Texas, so that I could drive to Del Rio, Texas, where I thought I would find the climbing species of Ephedra, E. pedunculata (a wonderful specimen was, in fact, growing up a telephone pole within the city limits there and I harvested it).  You don’t find many gnetalean woods in wood collections. There are a only a few samples of Gnetum in xylaria—and about a third of them aren’t Gnetum!  Most species of Gnetum are lianas, and when people have collected wood of lianas, they see flowers or fruits up there in the trees, and the stems of the woody vining stems close to the ground, and they collect the latter hoping they represent the former.  But often, the wood near the ground isn’t connected to what they see twining up there in the trees.  You can see that for various reasons, collecting my own materials has been vital to my progress in wood anatomy.

   Included in the discoveries were near-vessellessness in high-altitude species of Ephedra [ PDF ].  I began using a scanning electron microscope at about that time, and therefore, I was able to demonstrate the occurrence of helical thickenings in vessels of many New World ephedras [ PDF ].  In both Old World and New World species of Ephedra, tiny calcium oxalate crystals occur among tracheids and among ray cells, a discovery I made thanks to the SEM [ PDF ].  Tori could be shown on vessel pits of Gnetum by means of SEM.  In Welwitschia [ PDF ], occurrence of true secondary xylem  with rays and fascicular areas in it were demonstrated.  The fact that increments of the wood are produced by successive cambia may have led people to think that Welwitschia had bundles rather than increments of secondary xylem.   The use of a softening technique combined with paraffin sectioning permitted some amazing preparations [ PDF ]. 

   But most importantly, I saw that wood of Gnetales is essentially conifer wood in which vessels had developed, and it has nothing to do with angiosperms.  I was the only one at the day-long symposium on Gnetales at San Diego who presented the conifer concept for Gnetales, and stated it clearly, citing also some early DNA work by Hasebe et al.  The remaining participants seemed determined to show or agreeable to the idea that Gnetales were ancestors of angiosperms.  Subsequent DNA work has shown that Gnetales are definitely either within conifers or adjacent to conifers.   But aside from being ahead of my time in affirming the correct position of Gnetales, I learned an enormous amount about the woods of the group, summarized in a paper published in a special issue of International Journal of Plant Sciences [ PDF ].       

   The similarities among the three genera of Gnetales are at least as conspicuous as the differences with respect to wood anatomy.  At one time, the three genera were regarded as not at all close.
     I wrote two main papers on the wood anatomy of Ephedra.  The one on New World species, came first because I had material of more species, since I collected many of them myself.  Material of the Old World species was more difficult to obtain and took longer to assemble, so the paper on those species came later.  At the time I wrote the two papers, the prevailing infrageneric classification did not correspond to a divide between New World and the Old World species, but molecular work has since shown that this geographical divide corresponds to the main phylogenetic division within the genus. 
   The story of Gnetales here can be told more by pictures than by words because I took color pictures of the three genera during my field work, and I took color photomicrographs in preparation for the talk in San Diego.

   In 2012, I published a paper in which I reviewed the wood physiology of Gnetales. Because the three genera of this ancient group of conifers range from wet tropical forest to freezing high-elevations of the Andes and Himalayas, the nature of their wood in relation to wood physiology is important to know. Some botanists, looking at wood of various plants, have thought that wood formulas of particular plant groups have doomed them to restriction to particular ecological sites. Gnetales are rarely a prominent element in the world flora, some why are they rare in terms of the total area they occupy? As a group, they are older than flowering plants. They have had more time to occupy many kinds of habitats. In reviewing wood features, I found that there are essentially no features of angiosperm wood that are not also found in Gnetales. So wood structure is not limiting Gnetales to particular ecological niches. What is? The obvious explanation is that they have a longer life cycle, as gymnosperms, and therefore replace themselves more slowly in a habitat than angiosperms do. There are no annual Gnetales, no short-lived perennials, and they grow for years before flowering. In studies of desert areas that include Ephedra after burns, researchers have found that Ephedra seedlings are slow to revegetate burned areas compared to flowering plants like Artemisia.