The Tree (15 page)

A O
NE-OFF FROM
S
OUTHERN
J
APAN
: F
AMILY
S
CIADOPITYACEAE

Here is another very small family—indeed, with just one species:
Sciadopitys verticillata.
It’s an evergreen conifer endemic to southern Japan, where it grows on steep slopes and ridges, sometimes in clumps and sometimes mixed with broad-leaved trees. It competes because it can tolerate such poor soils.

The Sciadopityaceae are an ancient group; fossils that possibly belong to this family are known from the Upper Triassic, around 220 million years ago, before the dinosaurs got fully into their stride. Traditionally it has been included in the Taxodiaceae (which means it would now be in the Cupressaceae). But its needle-like leaves are not leaves at all—again, they turn out to be phyllodes. Also, studies of structure and of DNA support the notion that the remarkable
Sciadopitys
indeed belongs in a family of its own.

T
HE
Y
EWS
: F
AMILY
T
AXACEAE

In the family Taxaceae are five genera, with twenty-three species between them. Mostly these are in the Northern Hemisphere, but there are a few in the south, including—inevitably, it seems—in New Caledonia. The only widespread genus is
Taxus,
the yew, with ten species found throughout North America and down into Honduras, and in Eurasia down into the Southern Hemisphere in Malaysia and Indonesia. The British are used to the common yew
(Taxus baccata)
lowering gloomily in dank churchyards. Yet it lives in the tropics too, albeit confined to mountainsides.
Torreya
is less widespread, but also occurs in Asia as well as North America. Trees of the Taxaceae are slow-growing and long-lived—hence the somewhat speculative though not to say outlandish claim in the preface that Pontius Pilate might have dozed beneath one that is still growing in Scotland. Members of the Taxaceae seem doleful by nature; adapted to the shady understory in coniferous or mixed forests. The fruit-like arils, of various colors, are dispersed by birds. Like the podocarps, the group as a whole has not been studied enough. In the past, because of the peculiar and much-reduced female cones, yews were thought to form a group on their own, separate from the conifers. But closer study suggests that their cones have evolved from more complex seed cones, and this, plus evidence from DNA, confirms yews as bona fide conifers.

The wood of yew is valuable in many ways. In particular, it provided the medieval English with their longbows.

I
T WOULD BE GOOD
to devote this entire book to conifers. They are so various and wondrous and in many ways, in many contexts, they have turned the course of human history. Yet they are at least matched and in some ways far outstripped by the flowering trees, which will occupy the next five chapters.

Beautiful but simple: is magnolia the most ancient flowering tree?

6

Trees with Flowers: Magnolias and Other Primitives

T
HERE IS MORE TO FLOWERS
than meets the eye. Of course they may be beautiful. They are also, beautiful or not, superb essays in engineering, wonderfully efficient first at achieving fertilization, and then at producing and dispersing their seeds. The archetypal flower is supported by a circle of sepals, usually green, that make the calyx; inside that is a circle of petals, which between them form the corolla; and in the middle are the sexual parts—the male stamens, with the anthers at the tips, containing pollen, and the female carpels. Each carpel has an ovary with its ovule (or ovules) inside, a projecting style, and is tipped by the stigma, which receives the pollen. Immediately we see a key difference between angiosperms and all other seed plants. In the angiosperm the ovule is completely enclosed within the ovary, and the male gamete (reduced to a nucleus) is carried to it via a pollen tube that must burrow through the full length of the tissue of the style. In conifers and all other seed plants, the ovule is not completely enclosed. The pollen tubes do not have to burrow through living tissue. In cycads and ginkgoes, motile sperms do the fertilizing.

Then there is another key difference, one much less obvious. Uniquely, angiosperms practice “double fertilization.” As noted earlier, the pollen contains the male gamete—and other cells too. So, too, does the ovule—a true egg cell, and also subsidiary cells. In all but the most primitive angiosperms, the male gamete fuses with the egg to form a new embryo, as in all other organisms that practice sex. But a second cell in the pollen fuses with two of the cells in the ovule to form a combined cell with three sets of chromosomes; and this peculiar triploid cell then multiplies to form a food store, rich in carbohydrates, protein, and often fat, that surrounds the embryo. Double fertilization is a very neat trick—and unique to angiosperms.

The astounding ability of angiosperms to take so many different forms—mighty trees and creepers, tiny floating duckweeds, and everything in between—seems at first sight to have nothing to do with the innovations of flowers and seeds and fertilization. But it happens anyway. Perhaps the more sophisticated devices of sexual reproduction simply allow new ways of living.

Flowers are immensely variable. Some are huge and showy, others cryptic to the point of invisibility. Some have all the standard parts—sepals, petals, stamens, and carpels—but others have abandoned one or several of the basic components, and some incorporate various bracts (modified leaves) and other structures. In some the petals are green like sepals, and in others the sepals are colored like petals, and in some the two kinds of structure are more or less indistinguishable. Most flowers are hermaphroditic (like those of magnolias), while others are single sex; and some monoecious angiosperms keep both sexes of flowers on the same plant (like oaks), while other dioecious types have only one sex per plant (like holly). The sexual organs themselves take many different forms (as is also true of animals) and are key features for identification.

Many flowering plants are pollinated by various kinds of animals—not just insects but also bats, birds (such as hummingbirds), and sometimes mammals, from fruit bats to giraffes (or so it is said). Best known among the insect pollinators are bees, butterflies, and moths, but flies and beetles are important too. Beetles were probably the first insect pollinators of all—and probably developed the trick in association with cycads. The brash new angiosperms lured at least some of them away.

But many flowering plants, including most temperate trees, are pollinated by wind. Some—probably more than is yet appreciated—make some use both of animals and of wind. In a few, like the sea grasses, which flower underwater, and perhaps some plants of the mangroves, the pollen is conveyed by water. In general, the animal-pollinated flowers are showy and the wind-pollinated ones much less so. But flies, for example, hunt largely by scent and pollinate many very inconspicuous flowers (like those of the garden shrub
Fatsia,
a close relative of ivy), while wind-pollinated flowers often take the form of catkins, which can be very spectacular indeed.

In the plants that are considered to be most primitive, all the petals are much the same, all the sepals are much the same, and so on; and all the parts of the plant are separate. So the flower consists simply of repeated modules, arranged in spirals like the scales of a cone (although let me emphasize that flowers are not botanically related to cones). The whole is a kind of Kiddicraft flower. Magnolias and waterlilies are of this primitive type. Such flowers are radially symmetrical: symmetrical whichever way you look at them. In other flowers the different parts may be fused and extended in all kinds of shapes, and are commonly designed to attract very particular kinds of insects, and then to ensure that the visitors are well furnished with pollen. In such types (and many wind-pollinated types) the flowers are bilaterally symmetrical, like a face, or indeed are asymmetrical. Orchids are perhaps the most famous floral elaborators; but many other families have elaborate, specialist flowers too, including the pea family, Fabaceae (Leguminosae), which includes many of the world’s most significant trees. But primitive flowers can also trap insects in very clever ways. “Primitive,” after all, is not a pejorative term. It merely means “close to the ancestral state.” Daisies, incidentally, may
look
primitive (radially symmetrical, repeated parts). But in truth their flowers are complex inflorescences (collections of flowers), each of which is highly modified. So the daisy family, the Asteraceae (formerly called the Compositae), is commonly considered to be the most “advanced” of all. Perhaps unsurprisingly, though in truth for different reasons, the Orchidaceae and the Asteraceae both include a spectacular number of species—far more than any other family. Of the two, however, only the Asteraceae has trees—though not many and little to write home about. The family with the most trees (as well as many of the grandest) is the Fabaceae.

The fruits of flowering plants, containing the seeds, are also immensely variable. Some are big, bright, and fleshy and are intended to be spread by animals. Some (notably orchids) are tiny and windblown. Some windblown types are bigger but are fitted with wings. This is the case for many trees—such as the familiar sycamore and ashes, but perhaps most spectacularly in the dipterocarps, the great forest trees of Southeast Asia, which Southeast Asians, at least, consider to be the most important tropical trees of all. Other airborne seeds are fitted with cottony extensions that serve as parachutes. We all know dandelions and groundsel. In Yorkshire I have driven through blizzards of migrating thistle seeds that seemed to go on for miles. Cotton, too, of course (a relative of the hollyhock), produces its fiber as a way of spreading its seeds. Among trees, perhaps the most spectacular and significant parachute seeds are those of the various ceibas—their cottony kapok is used to stuff mattresses and parkas. Some fruits, notably the coconut and the giant
coco de mer,
are expressly designed to float.

So various are flowering plants, it has often been suggested that they cannot be a single group, not a true clade, with just one common ancestor. But they clearly are. All but the most primitive practice double fertilization, and that is so weird, and so complicated, that it surely could not have evolved more than once. So flowering plants, for all their marvelous variety,
are
a single invention, all derived from one common ancestor, which arose about 145 million years ago. Truly they are one of nature’s greatest inventions.

But it is not clear who the common ancestor was, or what it looked like. There are two ways to go about finding out. One is to look at the fossil record, to try to see what the very first flowering plants looked like. The other approach is to look at living plants, decide which are the most primitive, and then assume that the very first flowering plants must have looked roughly like the still-existing primitives. (The latter exercise is helped by applying cladistic principles to the DNA, and inferring which plant’s DNA is the most basic, with the most shared primitive features. But the details need not delay us.)

Both approaches are necessary, but both are riddled with traps. The fossil record is notoriously patchy (or “spotty,” as the palaeontologists tend to say). We are most unlikely to find the first organisms in a new lineage because, obviously, new lineages begin with just a few individuals—and rare organisms have scant chance of being fossilized, and then recovered. More generally, if we fail to find what we are looking for, that does not mean that our quarry did not exist. But at least if we do find something, that shows that it did exist. In reality, some of the earliest known angiosperm fossils are waterlilies, which seems to fit very nicely with expectation, since waterlilies have a simple kind of flower. But waterlilies are herbs. The problem with this is that the timber of flowering or broad-leaved trees closely resembles the timber of conifers. That structure is very complex, and is not likely to have evolved more than once. This and other evidence suggests that conifers and angiosperms shared a common ancestor. If flowering plants had first arisen as herbs, they would have to have reinvented timber that was very like that of conifers, and this seems most unlikely. So it seems that the very first flowering plants must have been trees—and thus for all their primitive flowers, waterlilies are highly evolved specialists, devoid of wood. They could not have been the first. The first angiosperm must have been a tree. Magnolias are reasonable candidates, but the oldest known magnolias are not quite old enough, and it is hard to imagine that the first ever angiosperm took quite such spectacular form.

Then again, among living angiosperms, two quite opposite kinds of flowers have been mooted as the most primitive. One is the big showy kind, as in waterlilies and magnolias. The other is a simple but small and modest kind, as found in pepper vines (the plants that produce peppercorns, not sweet or chili peppers). But the first ancestor of all the flowering plants
either
had magnolia-like flowers
or
pepper-like flowers. It could not have had both. If it had some compromise form, able to evolve either way, then we simply do not know what it was. The same kind of dilemma applies to fruits. Are the most primitive the small and scrawny kinds or the big and fleshy kinds, packed with seeds, like custard apples?

So the origin of flowering plants remains mysterious, but we do know that they are now immensely various, and their variety is reflected in their taxonomy. Roughly speaking, there seem to be around 300,000 species. Botanists differ in the ways they divide these species into families: recent published papers recognize anywhere between 387 and 589 families. The specialists who form the Angiosperm Phylogeny Group have, for the time being, plumped for 462. Only specialists can get their heads around so many, so it is useful to group the families into orders. Again, different taxonomists recognize different numbers of orders, but the authorities I am following in this chapter (whom I refer to collectively as “Judd”) divide them into 49, most of which contain trees. Forty-nine is still too many for most people to bother with, but most of the trees are contained within about 30 of the orders, and that is manageable.

One last word before the catalog begins—on the idea that was first mooted by John Ray in the seventeenth century, and was developed by Antoine-Laurent de Jussieu at the end of the eighteenth. Ray (you will remember from Chapter 2) divided flowering plants into those with narrow leaves and those with broad leaves; and Jussieu found that the narrow-leaved types were monocots (with a single cotyledon) while the broad-leaved types were dicots (with two cotyledons). So the distinction has stood for the better part of two hundred years. In rough-and-ready terms, the distinction still stands. But those interested in phylogeny—the true history of plants—as opposed merely to convenience must now make a serious modification. For it is clear that the first ever flowering plants were dicots. Some of those early, primitive types are still with us—including the magnolias, the waterlilies, and the peppers.

After some time, there arose from among the ranks of the primitives a new group that went off in novel directions. Some of these avant-garde types evolved into the monocots—which thus emerge as comparative latecomers. Some evolved into a quite new kind of dicot, now known collectively as the “eudicots” (“eu” meaning “good”). The monocots are a true clade—all of them derive from a common ancestor. The eudicots are also a true clade. It is also possible—indeed, quite likely—that the ancestor of the monocots was also the ancestor of the eudicots. In that case, the monocots and the eudicots together form a true clade.

But the primitives (magnolias, peppers, and so on) remain out on their own: not members of either of the two modern clades (although flowering plants as a whole form a true clade).

In effect, we now have three great groups of flowering plants. First there is a mixed bag of “primitive dicots”: not a clade, just the kinds that seem to have retained many of the main features of the first ancestor. Then there are the monocots, which are a true clade. Then there are the eudicots—another true clade, this time of derived dicots. The primitives are discussed in the rest of this chapter; the monocots have the next chapter to themselves; and the eudicots are spread over the three chapters after that (because they are too diverse to be accommodated comfortably in one).