- Email: [email protected]
Why do trees live so long?
Ageing Research Reviews 1 (2002) 653–671
Viewpoint
Why do trees live so long? Ronald M. Lanner∗ Institute of Forest Genetics, Pacific Southwest Research Station, USDA Forest Service, 2480 Carson Road, Placerville, CA 95667, USA Received 19 April 2002; accepted 20 April 2002
Abstract A long life multiplies a tree’s reproductive opportunities, thus increasing its fitness. Therefore, characteristics that prolong life should be naturally selected. Longevity in trees is achieved by means of numerous behaviors and characteristics, some of which are unique to trees. These include the retention of stem-cell-like meristematic cells after each growth cycle; the ability to replace non-vigorous, lost, or damaged organs, both above and below ground, in the presence or absence of trauma; a sectored vascular system that allows part of a tree to survive where a whole one cannot; formation of clones; a mechanical structure that can react to forces tending to de-optimize it; a hormonal control system that coordinates the above behaviors; and synthesis of defensive compounds. © 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Aging; Conifers; Evolution; Tree growth; Clonal growth
1. Introduction Trees are the exemplars of long-lived organisms. For sheer verifiable longevity, no other organism rivals that of the Great Basin bristlecone pine (Pinus longaeva), several of which exceed 4000 years of age (Schulman, 1958; Currey, 1965), and one of which has attained at least 4862 years (Lanner, 1999). But trees of numerous other species routinely live 1000 years or more, and a much greater number become centenarians (Kozlowski, 1971). Ancient Great Basin bristlecone pines show no inherent signs of senescence (Lanner and Connor, 2001): their potential life spans are limited by the activities of pests, the frequency and intensity of fires, and ultimately, how long it takes for the soil they are rooted in to erode away. Maximum life spans are fascinating to contemplate. But the question dealt with here, Why Do Trees Live So Long?, addresses the underlying idea that there must be something or things about trees that make them, and not some other class of organisms, ∗
Present address. 2651 Bedford Avenue, Placerville, CA 95667, USA. Tel.: +1-530-344-0255. E-mail address: [email protected] (R.M. Lanner).
1568-1637/02/$ – see front matter © 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 1 5 6 8 - 1 6 3 7 ( 0 2 ) 0 0 0 2 5 - 9
654
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
candidates for extremely long lives. So the purpose of this article is to isolate some factors and behaviors that are peculiar to trees, or even unique to them, that seem influential in making them potential elders and ancients among the world’s living things. In effect, we will be searching out the treeness of trees. A long life should be a selective advantage for a species that can remain reproductive into great old age, as is the case with trees, because it increases the likelihood of successful reproduction. In temperate-zone trees, reproductive maturity may be delayed for one to several decades after germination and establishment of the seedling; but a sufficiently long life creates the potential for many reproductive episodes. For example, a Rocky Mountain white fir (Abies concolor) (Fig. 1) that starts to produce seed at 40 years of age (Schopmeyer, 1974) lives to 350 years, and has a seed crop every 3 years has about 103 opportunities to encounter a fruitful seedbed when randomly dispersing its wind-borne propagules. At the extreme, a Great Basin bristlecone pine of exceptional age may cast seeds 2000 times or more during its full life. So if there are heritable traits that increase longevity, they should be selected.
Fig. 1. Most conifers, like this Rocky Mountain white fir, produce large cone crops at intervals of several years. Great Basin National Park, Nevada.
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
655
Schulman (1954) pointed out that the oldest Great Basin bristlecone pines grow under rigorous site conditions: short growth season, high winds, low temperatures, low precipitation, and often on barren rocky substrates where few other plants survive. As a result the generalization ‘adversity begets longevity’ has had wide currency. But the next oldest species, giant sequoia (Sequoiadendron giganteum) grows at lower elevations then bristlecone, in dense forest on warmer, moister sites, among very tall specimens of several associated species (Harvey et al., 1980). Other long-lived trees can be found under either set of conditions, so Schulman’s generalization is unpersuasive. While bristlecone’s longevity is largely due to escape from fire, and those of its predators that cannot tolerate harsh mountain climates, giant sequoia owes its longevity to pest-repellent secondary defensive compounds and the fire protection afforded by thick bark. Thus, the two most famous of oldsters, whose natural ranges are separated by just a few miles, follow divergent strategies for living a long time. But both are trees, and what they have in common is much greater than what sets them apart. A tree manages to live for many years by predictably repeating a pattern of behavior, year after year, that has been refined by natural selection. Listed below are some of those key behaviors that make long life possible.
2. Creating a structure that can be expanded First a tree must create a structure that can progressively be expanded, so it can hold itself rigidly erect above other vegetation, and bathe in sunlight. The following discussion describes briefly the major growth activities of a tree (Wilson, 1970; Kozlowski, 1971). The growth of bark from the cork cambium will not be considered here. It should be borne in mind that rates and timing of growth are controlled mainly by environmental factors, but the growth patterns described below are genetically determined. 2.1. Growth in height Trees get taller by adding new growth increments to the ends of their vertical shoots, or ‘leaders’. This is referred to as primary growth. The new growth originates from the division and enlargement of cells localized in apical meristems. In the temperate zones, most trees’ apical meristems are located at the summits of dormant buds and are protected by a cover of bud scales, until the bud axis expands in the spring. Tropical trees’ apical meristems are usually protected by a looser rosette of embryonic leaves. The dome-shaped apical meristem is the locus of formation for leaf primordia. As these primordia enlarge, growth just below the bud axis—in the subapical region—forces the apex upwards in a surge of growth that is at first slow, then rapid, and then slow until it ceases. Maturation of the new tissues is wavelike, beginning in the lower (therefore older) part of the new shoot, proceeding upwards into the upper (therefore younger) part. The apical meristem typically ends the growth cycle by forming a new bud for the next cycle. Formation of a bud never ‘uses up’ the meristematic tissue of the apical meristem, but in some species a floral structure may form, thus precluding further lengthening of that axis. The final act of maturation is the lignification of the shoot’s supportive tissues, making the shoot ‘woody’.
656
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
This is accomplished by the deposition of lignin inside the cellulose walls of their cells (Panshin and de Zeeuw, 1970). In the early stage of shoot elongation it is typical for the new leaves to receive photosynthate from storage tissues in the previous year’s growth of the tree, but as the season of growth progresses, the maturing leaves become more self-sufficient, and finally produce a surplus of carbohydrate which is translocated to new storage tissues. This is usually in the ray parenchyma or axial parenchyma of the new woody shoots. In conifers, the needle leaves also serve as storage organs (Kozlowski, 1971). Meanwhile, other shoots are growing outwards in the same way, increasing the spread and complexity of the tree’s crown. 2.2. Growth in diameter As a tree grows taller, its trunk must thicken, or it will collapse. Diameter growth of trunks and limbs is by the division and enlargement of cells of a lateral meristem, the vascular cambium, a process called secondary growth. The vascular cambium, or cambium, as it is usually known, is a multicell-thick sheath or jacket of undifferentiated meristematic cells found between the outermost annual ring and the inner bark. When activated, the cambium produces daughter cells to its inside that rapidly differentiate into xylem (wood) cells. In broadleaved trees (angiosperms), most of these become fibers, which have a support function; or vessel cells, which align to form the pipelines that conduct xylem sap. In conifers, which have simpler wood with fewer cell types, both support and xylem sap transport are combined in tracheids. These are long, narrow cells tapered at both ends and aligned longitudinally in the stem. Vessel cells and tracheids formed early in the season (earlywood) have thinner walls and larger diameters than those formed later (latewood). The contrast between latewood of 1 year and earlywood of the next year delineates the ‘annual ring’ in cross-sections of temperate-zone trees. Tropical trees have indistinct non-annual rings that reflect episodes of growth. Daughter cells that form at the cambium’s outer periphery differentiate as phloem, or inner-bark cells. Most of them become sieve cells or sieve tube members which specialize in the polar transport of photosynthate, growth substances, etc. formed in the tree’s leaves and distributed throughout the tree. Unlike xylem sap-conducting tracheids and vessels, which must be dead and emptied of their contents in order to function, sieve cells must be alive to function. Much less phloem than xylem is formed in a growth cycle. Phloem cells typically live only 1 year, and are crushed against the wall of bark by outward expansion of the xylem. New phloem is formed at the beginning of the next growth cycle. Though conductive xylem cells die in their first year, ray parenchyma and axial parenchyma associated with them live for several years, even decades (Connor and Lanner, 1990). As new growth rings are added, older ones back in the stem experience death of their ray parenchyma, deposition of phenolics, etc., and frequently, discoloration (Panshin and de Zeeuw, 1970). The outer wood is referred to as sapwood, the inner as heartwood. Heartwood is usually considered to be biologically inert. 2.3. Root growth The primary growth of roots differs from that of stems by not being from buds, but from exposed apical meristems. They usually grow mostly when shoot growth is quiescent.
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
657
Secondary growth resembles that of stems, though many fine roots die shortly after forming and never become woody. Symbiotic relationships with mycorrhiza-forming fungi or with nitrogen-fixing bacteria or actinomycetes, increase the efficiency and probably extend the lives of many tree roots. The long-term goal of a tree is to become large enough and live long enough to produce seed. In order to do that it follows a dual strategy of first capturing space by rapidly growing upwards and outwards in prematurity; and then exploiting that space by filling it with foliage-bearing branches (Edelin, 1977). Analogous territorial claims continue underground, through a ‘heterorhizic’ system of long roots (or pioneer roots) which invade new ground, and dense mats or fans of short roots, which exploit the new ground between them (Sutton, 1969).
3. Making the structure redundant A structure that has more components than it needs at one time is regarded here as having the property of redundancy. There are several ways in which a tree may neutralize damage or loss of vital parts in order to live another day; or even perpetuate its genotype by replicating it. 3.1. Epicormic branching and stump sprouting as a response to injury Trees form many buds that fail to elongate into shoots the next year. Instead, they remain dormant. Buds are formed in the angle where a leaf is joined to a stem, so dormant buds are at risk of being overwhelmed by the radial growth of the stem, and embedded within it. However, the vascular trace that supplies sap to the bud grows a little each year, and allows the bud to remain alive while staying at the outer periphery of each year’s new sapwood. The buds’ growth is held in check by inhibitory growth substances produced in the leaves and moved downwards in the phloem. If the crown is damaged in a way that interrupts the flow of the growth substances, the dormant buds are released, and form new shoots that emerge from the stem surface. Termed epicormic branches, these shoots are a common response to pruning, wind breakage, defoliation, sudden exposure to bright sunlight, and other mechanical injuries to tree crowns. Epicormic branches are virtually ubiquitous in broadleaved trees, and have been considered relatively uncommon in conifers. During its sojourn beneath the bark, the dormant bud (or apical meristem, if the bud is sufficiently undeveloped) may branch profusely, forming a large number of potential shoots. This can produce a mass of new shoots very rapidly upon injury. In a few pines, buds long held dormant beneath the bark of the trunk sprout after fire scorches the tree’s crown. These include Canary Islands pine (P. canariensis), pond pine (P. serotina) (Fig. 2), pitch pine (P. rigida), and Chihuahua pine (P. leiophylla). A type of epicormic branch unique to pines is the interfoliar shoot that develops from the apical meristems within the needle bundle, or bundles, just proximal to the site of injury. Cooperrider (1938) described in detail the occurrence of this phenomenon in ponderosa pines (P. ponderosa) in the southwestern US that had been browsed by cattle. It is also a common response to defoliation injury.
658
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
Fig. 2. Pond pine is one of several pines that produce epicormic branches on their trunks following fire damage. Dismal Swamp, Virginia.
Another example of epicormics specialized to the needs of a particular tree species is the sequestering of innumerable buds on the upper surface of branches of the California endemic bigcone-spruce (Pseudotsuga macrocarpa). This species grows in highly combustible vegetation communities. When fire scorch kills its primary branches, they are soon clothed by a dense growth of newly sprouted foliage. This habit has been noted even in trees up to 300 years of age (Lanner, 1999) (Fig. 3). If an entire tree is badly injured, shoots of epicormic origin may emerge from the base of the trunk. Or if a tree is cut down, such shoots (called ‘stump sprouts’) may emerge from the cut stump. This is the basis of an ancient forestry practice, called ‘coppicing’, which consists of periodically felling trees for small wood products, and encouraging the growth of new coppice shoots for another crop. If the branches are cut off, instead of the whole tree, new branches sprout just below the point of shearing, a practice termed ‘pollarding’. Pollarding has been used as a means of growing firewood. Even when coppicing is not intended, it can occur on a large scale, as in the northern hardwood forests of the northeastern US, where heavy cutting of oaks, maples, and
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
659
Fig. 3. Bigcone spruce produces great numbers of epicormic branches following scorching by brush fires. Los Padres National Forest, California.
other trees that sprout vigorously, produced an extensive second-growth forest in the 19th century. A variant of the stump sprouts so common in broadleaved trees, but less so in conifers, is the formation of ‘fairy rings’ of fast-growing redwood (Sequoia sempervirens) sprouts surrounding the base of a dead tree, or the stump of one that has been cut (Barbour et al., 2001). It can hardly be doubted that epicormic branches prolong the life of the trees that produce them following injury; and that stump sprouting prolongs the life of its genotype. The question of prolongation of life via cloning will be considered below. 3.2. Epicormic branching as an ontogenetic shift While epicormic branching in broadleaved trees is usually a response to injury, it has been shown in recent years that some conifers form epicormics in the absence of trauma.
660
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
Fig. 4. A stand of old-growth western larch in which original primary branches have largely been replaced by epicormic branches. See especially the three trees left of center. Glacier National Park, Montana.
The primary branches of western larch (Larix occidentalis)—those formed on the trunk— become a liability that can limit tree age. As a branch ages it gets longer, more branched, and heavier. It is additionally burdened by the weight of rainwater, lichen growth, snow, bird nests, witches’ brooms, and wind torque. It can act like a sail in the wind, increasing susceptibility to windthrow, if it is well-anchored; or breaking off and leaving a gaping wound if not. But as such a branch ages, clusters of new green epicormics sprout from its base. In a typical 300–500-year-old western larch, two thirds of the crown may consist of long, dead primary limbs reaching out starkly, while masses of epicormics hug the trunk (Fig. 4). In the upper crown, the younger, smaller, more productive primary branches still prevail; but for how much longer is not known. Eventually, all the tree’s foliage may be borne on epicormics. This shift in the ontogeny of larch branches—what I refer to as an ontogenetic shift—allows the larch to maintain a crown of vigorous branches, surely extending its life for many years (Lanner, 1996). Several years before I reported this habit in western larch, Bryan and Lanner (1981) described it in Rocky Mountain Douglas-fir and coast Douglas-fir (Pseudotsuga menziesii vars. glauca and menziesii). We pointed out that there was apparently a ‘genotypically based program of crown maintenance that materially prolongs the life of the tree, and that this program requires no exogenous triggering mechanism’. Both in the Douglas-firs and the western larches, there were no indications of trauma having stimulated the epicormics, as is the case in broadleaved trees. Further study of old-growth coast Douglas-fir (Ishii and Ford, 2001) has confirmed that ‘epicormic shoot production maintains shoots and foliage of old P. menziesii trees after height growth and crown expansion have stopped, and may contribute to prolonging tree longevity’.
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
661
I have observed this phenomenon also in European larch (L. decidua), subalpine larch (L. lyallii), giant sequoia (Sequoiadendron giganteum), coast redwood, Sitka spruce (Picea sitchensis), and western hemlock (Tsuga heterophylla). There is in the pines an analogous situation to this, in which interfoliar branches are formed from within needle bundles without being stimulated by trauma. This occurs routinely in the bristlecone and foxtail pines of Pinus subsection Balfourianae (Connor and Lanner, 1987). 3.3. Sectoriality of the vascular system A tree’s woody skeleton can be viewed as a bundle of pipes gathered in the middle into a trunk, and radiating at either end into the root and branch systems (Shinozaki et al., 1964). One can further envision that the transport of water from the soil into a specified branch takes place only or mainly or largely in the trunk pipe or pipes that connect that branch to ‘its’ root pipe or pipes. In other words, there is a minimum of tangential spreading from one flow-way into another. If that were the case, the tree would be a composite of several hydraulically independent sectors. There is evidence that in some trees sectoriality of the vascular system does occur. For example, Rudinsky and Vité (1959) injected dyes into the sapstream of several conifer species. By examining the occurrence of the dye in different parts of the stem cross-section at different heights, they inferred five patterns of water ascent. Among these were ‘sectorial, winding ascent’, and ‘sectorial, straight ascent’. The spirality of ascent was correlated with the alignment of xylem cells. More recently, Larson et al. (1994) have shown that in very old northern white-cedars (Thuja occidentalis) that have been dye-injected, pathways of water transport appear to be sectored. Sectoriality of the vascular system is also indicated by field observations of the so-called ‘strip-bark’ habit of Great Basin bristlecone pine (Lanner, 1984, 1999) and ‘partial cambial mortality’ of Rocky Mountain bristlecone pine (P. aristata) (Schauer et al., 2001). Among very old trees one commonly observes trunks of bare wood that still retain only a narrow strip of bark. Beneath the bark, a live cambium produces annually a narrow strip of xylem and phloem. These strips can often be seen to extend downwards onto a major exposed root, and to wind around it into the soil. The bark strip ascends the trunk onto a branch containing some of the meager foliage still remaining on the tree (Fig. 5). The death of the major roots that no longer support live wood and bark has been attributed to drying following exposure through soil erosion, in the case of Great Basin bristlecone pine (Lanner, 1984, 1999), and perhaps to wind action in the case of Rocky Mountain bristlecone pine (Schauer et al., 2001). In the context of this discussion, it would appear that sectoriality of the vascular system of these trees is sufficient to prolong a tree’s old age by allowing it to keep alive ever-diminishing portions of its structure. 3.4. Root system replacement Under unusual circumstances, whole root systems of trees can be replaced. A well-known example is in coast redwood (Barbour et al., 2001). Trees growing on flood plains are
662
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
Fig. 5. A 1500-year-old Great Basin bristlecone pine, mostly dead due to exposure of its roots by soil erosion, with one surviving root providing water to a few branches via a narrow surviving strip of living wood and bark. Dixie National Forest, Utah.
susceptible to anaerobic soil conditions when their root systems are buried in deep sediments. Temporary relief is brought by upwards growth of new root tips. Permanent adjustment to these conditions comes about when an entirely new system of lateral roots forms within the sediment layer. Repeated floods over centuries can result in several such tiers of roots. The writer has observed an analogous situation in Hawaii Volcanoes National Park, where ohia lehua trees (Metrosideros polymorpha) partially buried in volcanic ash have formed new root systems in the upper ash layer which replace the buried root system. In both these cases, it can be safely inferred that the new roots aided tree longevity. 3.5. Clone formation Many trees can propagate vegetatively, forming small to large clones whose individual ‘ramets’ can substantially outlive the original mother tree. Griggs (1938) has described
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
663
how lower limbs of subalpine fir (Abies lasiocarpa) can ‘layer’, that is, form roots when in contact with moist soil, sometimes forming a circle of rooted branches around the parent tree. These layers can turn their tips upward, growing like saplings. Eventually, the branch connection with the parent tree is severed by decay, and the layers are physiologically independent. When the parent tree finally dies, the much younger layers prolong the life of the genotype, forming hollow-centered ‘timber atolls’ around the site of the dead parent tree. This also occurs in high-elevation Engelmann spruce (Picea engelmannii) (Wardle, 1968). An interesting variation of layering occurs in northern Quebec, with black spruce (P. mariana). Payette and Delwaide (1994) have described layered individuals at the extreme of the species’ range of occurrence that have persisted clonally during a period unfavorable to sexual reproduction, yet are still producing cones. A variant of layering occurs when a fallen tree retains roots in the soil and stays alive, while some of its branches grow upwards like trunks; and the trunk forms roots beneath the branch bases. Eventually, the fallen tree dies and decays, but its rooted branches persist as individuals. This has been described in Nepal alder (Alnus nepalensis) on the Island of Hawaii (Lanner, 1964). Perhaps the ultimate extreme of redundancy provided through vegetative propagation is the formation of large clones by root-sprouting. Nowhere is this more obvious than in quaking aspen (Populus tremuloides) in the Rocky Mountains (Lanner, 1984) (Fig. 6). Clones are formed by the sprouting of stems from the root systems of aspens that at some time in the past became established from seed. Over the years, these groves of genetically identical trees spread, as each ramet grows its own roots, and produces further ramets. Clones of thousands of interconnected stems can occupy 100 acres or more, and slowly migrate
Fig. 6. A clonal group of quaking aspen trees that have arisen as root sprouts from an interconnected root system. Note the similarity in branching angle and bark color of the stems. Wasatch-Cache National Forest, Utah.
664
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
over the landscape, as they grow at one edge while perhaps dying out at another. Minor morphological or phenological differences between clones allow easy identification of each clone’s boundaries. Throughout the Rocky Mountains it is difficult to find seedling aspens, as regeneration from root sprouts is highly predominant. It is widely believed that aspen has survived in the Rockies since the ice age only because of its clonal habit (Cottam, 1954), because present conditions seldom allow for seedling survival. However, after the catastrophic forest fires of 1988 in Yellowstone National Park, seedlings were reported to be abundant. Many other trees and shrubs form root-sprout clones, but seldom on the scale of quaking aspen. If genotypes whose seedling requirements are temporarily unmet in a changing environment can form clones, they may be able to survive until conditions permit the resumption of sexual reproduction. Thus, in effect, cloning is a prolongation of the longevity of the genotype.
Fig. 7. The trunk of a tree develops as a ‘beam of uniform resistance’, thinner at the top and thicker at the base to resist stresses without waste of material. Wind-flexing stimulates cambial activity, thus, open-grown trees like this ponderosa pine have thicker, more tapered trunks than trees growing in dense stands. Dixie National Forest, Utah.
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
665
4. ‘Optimizing’ the structure Trees must obey physical laws. They are subject to constant gravitational force, and are impacted by wind, snow loads, soil creep, earthquakes, and other instabilities. Fortunately, they have evolved means of defeating some of the forces that threaten to topple them. Mattheck (1991) has referred to these means as adaptive growth that optimizes the structure by evening load distribution. A familiar example of load optimization is the taper of long conifer trunks, due to more wood being grown in the lower trunk where stress is greatest, and the least wood at the treetop. In this way, the trunk, a beam of uniform resistance (Büsgen and Münch, 1929), is as strong as it needs to be, and as economical in the use of its nutrients as it can be (Fig. 7). Mattheck (1991) places under the rubric of minimizing external loading such diverse tree growth responses as the bowing upwards of a leaning trunk, the upright growth of a limb below the point where a leading shoot has broken off (Fig. 8), apical dominance of leading shoots over lateral shoots in order to prevent the latter from encroaching on the former, and
Fig. 8. A limb of this limber pine swept upwards many years ago in response to the death of the tree’s leader. Humboldt National Forest, Nevada.
666
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
Fig. 9. Recumbent stems of Mugo pine are swept upwards by the pushing force of compression wood on their lower surfaces. Swiss Alps.
growth of stems directed towards a light source. Wounds that create ‘notch stresses’ are viewed as having their healing regulated by the mechanical stresses manifested as ‘force fields’. The cases of leaning stems and upsweeping branches are examples of corrections in tree or branch orientation by means of ‘reaction wood’. The variant of reaction wood known as compression wood forms on the underside of conifer stems and exerts a pushing force that tends to raise or uphold the stem (Figs. 9 and 10). Compression wood cells are distinctive, and form dark, crescent-shaped peripheries of the growth rings in which they occur. The other variant, ‘tension wood’, forms on the upper side of leaning angiosperm stems and branches, and is also anatomically distinct. It exerts a pulling force that tends to raise or uphold the stem in which it forms (Kozlowski, 1971). Mattheck (1991) views much of adaptive growth as an effort to rearrange the centre of gravity of a disturbed tree above the rootstock. He points out that much of this adaptation is accomplished by the variability in how new wood increments are distributed on the tree.
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
667
Fig. 10. A blue spruce that nearly fell into the Snake River when the bank collapsed two years earlier has already started to turn its tip upwards, due to the force exerted by compression wood, and phototropic growth of the new shoots. Grand Teton National Park, Wyoming.
Mattheck’s views comprise a modern mechanical engineering perspective on the mechanical design of trees, and deserve careful attention by biologists.
5. Hormonal coordination of growth activities It is obvious that the tree depicted here is not a mere stick of inanimate wood stuck into the ground. Rather, it is an exquisitely fine-tuned and reactive organism that contains within itself the means to regulate its own shape and size. For many years it has been felt that the plant growth hormone auxin, or indole-3-acetic acid (IAA) is responsible for a wide range of regulatory processes in trees, as well as in other plants. Experimental research has succeeded in identifying numerous such effects, but others have proven complex and difficult to demonstrate in a definitive fashion. For example, apical dominance effects have been largely attributed to the action of auxin, but experimental results
668
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
Fig. 11. The loss of apical control by the upper parts of a fallen but still rooted Fremont cottonwood has resulted in the vigorous vertical growth of its lateral branches. Green River, Utah.
have implicated the possible involvement of gibberelins, ethylene, and cytokinins as well (Wilson, 2000). Auxin moves in polar fashion down the tree from sites in the foliage where most of it is produced (Uggla, 1998). So it is not unexpected that apical control of shoots by shoots higher in the crown has been shown due largely to auxin (Wilson, 2000). And the obverse of this relationship, the release of inhibited buds when crown damage interrupts the polar transport of auxin, is a common explanation for epicormics and stump sprouts. This same effect is also the classical explanation for the release of root sprouts that results in widespread clones like those of quaking aspen (Lanner, 1984), and the upright growth of lateral branches on fallen trees (Fig. 11). Auxin has also been implicated in the formation of compression wood in conifers (Timell, 1986). Several studies have shown increased IAA concentrations on the lower sides of tilted stems, correlating with the development of compression wood at those sites; or that application of the auxin-antagonist TIBA prevents the formation of compression wood (Funada et al., 1990). Perhaps the most important role of auxin in trees is its function as a ‘long distance positional signal’ emanating from the tree’s crown and being received throughout the stem (Uggla et al., 1996; Uggla, 1998). In this process, auxin synthesized mainly in the new developing leaves moves in polar fashion down the phloem and affects the growth of cambial cells. It does not directly regulate rates of cell division, but rather the number of cells that are dividing. In this way it coordinates cambial growth with apical growth, an essential function for creating a tree of balanced form. A related need, which appears also to be served by auxin at least in part, is the hormonedirected transport of structural carbohydrates from storage areas into the new, growing shoots of the tree. The more new leaves and stem internodes between them that are unfolding,
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
669
the longer the new shoot will be; and the greater its need for structural material. It appears that auxin synthesized in the leaves moves downwards, and directs the upward movement of stored nutrients. Lanner and Connor (1988) showed that in ponderosa pine, elongation of the shoot was regulated by the needle bundles, presumed to be the source of auxin, rather than by the shoot apical meristem. Auxin may also participate in the control of photosynthetic rates due to strong sink activity (Tschaplinski and Blake, 1989; Wardlaw, 1990). It seems a reasonable conjecture that the rapid feedbacks in increased photosynthetic rates correlated with high sink demand (Dickson, 1991); or the decreases in photosynthetic rates following reduction of sink demand (Myers et al., 1999) are auxin-mediated at least in part. Though much research remains to be done to further elucidate the complex role of plant hormones in the regulation of tree growth and form, it is already apparent that growth regulators, and particularly IAA, are of primary importance.
6. Defending the structure from enemies This topic was thoroughly and creatively reviewed a number of years ago by Loehle (1988), so I will summarize some of the major points brought out in his study. Loehle analyzed energy partitioning between defensive investments and growth in trees, to test the hypothesis that greater longevity should require an increased energy investment in such defence measures as defensive chemicals and thick bark. He assumed there is a trade-off between using energy for rapid growth, and for defensive compounds. His data were from published information on 159 American tree species, both broadleaved and coniferous. Longevity of broadleaved trees was correlated with their investment in defences, but that of conifers was not. Longevity of conifers was, however, correlated with resistance to wood decay. Since his comparisons were cost-benefit analyses between tree species with differing life spans, they are not directly applicable to the question asked here. However, they do bring attention to the importance of defensive chemicals synthesized in trees to defend against threats to structural integrity posed by insect and fungal pests. Broadleaved trees of greater longevity spent a larger proportion of their lives in the reproductive phase than did short-lived species. Among conifers there was no such relationship, though there was some indication that in very long-lived conifers the attainment of sexual maturity might be contingent upon attaining a minimum size rather than a minimum age.
7. Conclusions Trees are long-lived organisms. It is advantageous for them to be so, because sustained presence in an environment gives them many more opportunities to produce and disperse propagules. Thus, it is assumed here that selection for relative longevity has the effect of increasing fitness. There are many behaviors and structural characteristics of trees that enable them to live longer than any other class of organisms. Those discussed here include annual modular growth from persistent meristems of stem-cell-like parenchyma cells, resulting in a new tree being created each year atop the tissues of the old one; redundancy of branch,
670
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
trunk, and root outgrowth either in the presence or absence of trauma, resulting in large-scale replacement of dead or worn-out organs; a sectored vascular system that postpones death; formation of clones whose ramets are individually short-lived, but in the aggregate are potentially immortal; adaptive growth that optimizes the mechanical structure of the tree, allowing it to regain verticality, use material sparingly, and resist mechanical forces; with all of these behaviors controlled and coordinated, it appears, by hormonal means; and the synthesis of defensive compounds to defeat insect and fungal enemies. Not all trees possess all the above behaviors, though some are common to all. And some life-prolonging characteristics have been omitted from this discussion due to time and space constraints. These include, but are not limited to, the compartmentalization of decay in trees, whereby rots are confined to tissues that pre-date the wound opening the entry court; the healing over of injuries by periderms and callus tissue; and the incorporation through natural grafting, of the root systems of suppressed trees into those of dominants in the stand. Since all these diverse characteristics found among trees make their own contributions to longevity, it would seem fruitless to try to ascribe tree longevity to a single gene or gene complex. What makes a tree long-lived is simply being a tree. References Barbour, M., Lydon, S., Borchert, M., Popper, M., Whitworth, B., Evarts, J., 2001. Coast Redwood, A Natural and Cultural History. Cachuma Press, Los Olivos. Bryan, J.A., Lanner, R.M., 1981. Epicormic branching in Rocky Mountain Douglas-fir. Can. J. For. Res. 11, 190–199. Büsgen, M., Münch, E., 1929. The Structure And Life of Forest Trees, 3rd Edition. Chapman & Hall, London. Connor, K.F., Lanner, R.M., 1987. The architectural significance of interfoliar branches in Pinus subsection Balfourianae. Can. J. For. Res. 17, 269–272. Connor, K.F., Lanner, R.M., 1990. Effects of tree age on secondary xylem and phloem anatomy in stems of Great Basin bristlecone pine (Pinus longaeva). Am. J. Bot. 77, 1070–1077. Cooperrider, C.K., 1938. Recovery processes of ponderosa pine reproduction following injury to young annual growth. Plant. Phys. 13, 5–27. Cottam, W., 1954. Prevernal leafing of aspen in Utah mountains. J. Arnold Arbor. 35, 239–248. Currey, D.R., 1965. An ancient bristlecone pine stand in eastern Nevada. Ecology 46, 564–566. Dickson, R.E., 1991. Assimilate distribution and storage. In: Raghavendra, A.S. (Ed.), Physiology of Trees. Wiley, New York, pp. 51–85. Edelin, C., 1977. Images de l’Architecture des Conifères. Thesis, Univ. des Sci. et Tech. Du Languedoc, Montpelier. Funada, R., Mizukami, E., Kubo, T., Fushitani, M., Sugiyama, T., 1990. Distribution of indole-3-acetic acid and compression wood formation in the stems of inclined Cryptomeria japonica. Holzforschung 44, 331–334. Griggs, R.F., 1938. Timberlines in the northern Rocky Mountains. Ecology 19, 548–564. Harvey, H.T., Shellhammer, H.S., Stecker, R.E., 1980. Giant Sequoia Ecology, Fire and Reproduction. Sci. Monogr. Ser. No. 12, National Park Service, Washington, D.C. Ishii, H., Ford, E.D., 2001. The role of epicormic shoot production in maintaining foliage in old Pseudotsuga menziesii (Douglas-fir) trees. Can. J. Bot. 79, 251–264. Kozlowski, T.T., 1971. Growth and Development of Trees, Vols. 1/2. Academic Press, New York. Lanner, R.M., 1964. Clones of Nepal alder in Hawaii. J. Forestry 62, 636–637. Lanner, R.M., 1984. Trees of the Great Basin. Nevada University Press, Reno. Lanner, R.M., 1996.The role of epicormic branches in the life history of western larch. In: Schmidt, W.C., McDonald, K.J., (Eds.), Proceedings of an International Symposium on Ecology and Management of Larix Forests: A Look Ahead, 5–9 October 1992, Whitefish, Montana. USDA Forest Service Intermountain Research Station Gen. Tech. Rep. 319, 323–326.
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
671
Lanner, R.M., 1999. Conifers of California. Cachuma Press, Los Olivos. Lanner, R.M., Connor, K.F., 1988. Control of shoot elongation in ponderosa pine: relative roles of apical and axillary meristems. Tree Physiol. 4, 233–243. Lanner, R.M., Connor, K.F., 2001. Does bristlecone pine senesce? Exp. Gerontol. 36, 675–685. Larson, D.W., Doubt, J., Matthes-Sears, U., 1994. Radially sectored hydraulic pathways in the xylem of Thuja occidentalis as revealed by the use of dyes. Intl. J. Plant Sci. 155, 569–582. Loehle, C., 1988. Tree life history strategies: the role of defences. Can. J. For. Res. 18, 209–222. Mattheck, C., 1991. Trees, the Mechanical Design. Springer, Berlin. Panshin, A.J., de Zeeuw, C., 1970. Textbook of Wood Technology, 3rd Edition, Vol. 1. McGraw-Hill, New York. Payette, S., Delwaide, A., 1994. Growth of black spruce at its northern limit in arctic Quebec, Canada. Arc. Alp. Res. 26, 174–179. Rudinsky, J.A., Vité, J.P., 1959. Certain ecological and phylogenetic aspects of the pattern of water circulation in conifers. Forest Sci. 5, 259–266. Schauer, A.J., Schoettle, A.W., Boyce, R.L., 2001. Partial cambial mortality in high-elevation Pinus aristata (Pinaceae). Am. J. Bot. 88, 646–652. Schopmeyer, C.S., 1974. Seeds of Woody Plants in the US. Agriculture Handbook no. 450, USDA Forest Service, Washington, D.C. Schulman, E., 1954. Longevity under adversity in conifers. Science 119, 396–399. Schulman, E., 1958. Bristlecone pine, oldest known living thing. Natl. Geogr. Mag. 113, 355–372. Shinozaki, K., Yoda, K., Hozumi, K., Kira, T., 1964. A quantitative analysis of plant form—the pipe model theory. I. Basic analyses. Jpn. J. Ecol. 14, 97–105. Sutton, R.F., 1969. Form and Development of Root Systems. Tech. Comm. No. 7, Commonwealth Forestry Bureau, Commonwealth Agric. Bur., Farnham Royal. Timell, T.E., 1986. Compression Wood in Gymnosperms. Springer, Berlin. Tschaplinski, T.J., Blake, T.J., 1989. The role of sink demand in carbon partitioning and photosynthetic reinvigoration following shoot decapitation. Physiologia Plant 75, 166–173. Uggla, C., 1998. New Perspectives on the Role of Auxin in Wood Formation. Acta Univ. Agric. Sueciae, Silvestria 58, Umeå. Uggla, C., Moritz, T., Sandberg, G., Sundberg, B., 1996. Auxin as a positional signal in pattern formation in plants. Proc. Natl. Acad. Sci. U.S.A. 93, 9282–9286. Wardlaw, I.F., 1990. The control of carbon partitioning in plants. New Phytol. 116, 341–381. Wardle, P., 1968. Engelmann spruce (Picea engelmannii Engelm.) at its upper limits on the Front Range, Colorado. Ecology 49, 483–495. Wilson, B.F., 1970. The Growing Tree. Massachusetts University Press, Amherst. Wilson, B.F., 2000. Apical control of branch growth. Am. J. Bot. 87, 601–607.
Viewpoint
Why do trees live so long? Ronald M. Lanner∗ Institute of Forest Genetics, Pacific Southwest Research Station, USDA Forest Service, 2480 Carson Road, Placerville, CA 95667, USA Received 19 April 2002; accepted 20 April 2002
Abstract A long life multiplies a tree’s reproductive opportunities, thus increasing its fitness. Therefore, characteristics that prolong life should be naturally selected. Longevity in trees is achieved by means of numerous behaviors and characteristics, some of which are unique to trees. These include the retention of stem-cell-like meristematic cells after each growth cycle; the ability to replace non-vigorous, lost, or damaged organs, both above and below ground, in the presence or absence of trauma; a sectored vascular system that allows part of a tree to survive where a whole one cannot; formation of clones; a mechanical structure that can react to forces tending to de-optimize it; a hormonal control system that coordinates the above behaviors; and synthesis of defensive compounds. © 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Aging; Conifers; Evolution; Tree growth; Clonal growth
1. Introduction Trees are the exemplars of long-lived organisms. For sheer verifiable longevity, no other organism rivals that of the Great Basin bristlecone pine (Pinus longaeva), several of which exceed 4000 years of age (Schulman, 1958; Currey, 1965), and one of which has attained at least 4862 years (Lanner, 1999). But trees of numerous other species routinely live 1000 years or more, and a much greater number become centenarians (Kozlowski, 1971). Ancient Great Basin bristlecone pines show no inherent signs of senescence (Lanner and Connor, 2001): their potential life spans are limited by the activities of pests, the frequency and intensity of fires, and ultimately, how long it takes for the soil they are rooted in to erode away. Maximum life spans are fascinating to contemplate. But the question dealt with here, Why Do Trees Live So Long?, addresses the underlying idea that there must be something or things about trees that make them, and not some other class of organisms, ∗
Present address. 2651 Bedford Avenue, Placerville, CA 95667, USA. Tel.: +1-530-344-0255. E-mail address: [email protected] (R.M. Lanner).
1568-1637/02/$ – see front matter © 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 1 5 6 8 - 1 6 3 7 ( 0 2 ) 0 0 0 2 5 - 9
654
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
candidates for extremely long lives. So the purpose of this article is to isolate some factors and behaviors that are peculiar to trees, or even unique to them, that seem influential in making them potential elders and ancients among the world’s living things. In effect, we will be searching out the treeness of trees. A long life should be a selective advantage for a species that can remain reproductive into great old age, as is the case with trees, because it increases the likelihood of successful reproduction. In temperate-zone trees, reproductive maturity may be delayed for one to several decades after germination and establishment of the seedling; but a sufficiently long life creates the potential for many reproductive episodes. For example, a Rocky Mountain white fir (Abies concolor) (Fig. 1) that starts to produce seed at 40 years of age (Schopmeyer, 1974) lives to 350 years, and has a seed crop every 3 years has about 103 opportunities to encounter a fruitful seedbed when randomly dispersing its wind-borne propagules. At the extreme, a Great Basin bristlecone pine of exceptional age may cast seeds 2000 times or more during its full life. So if there are heritable traits that increase longevity, they should be selected.
Fig. 1. Most conifers, like this Rocky Mountain white fir, produce large cone crops at intervals of several years. Great Basin National Park, Nevada.
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
655
Schulman (1954) pointed out that the oldest Great Basin bristlecone pines grow under rigorous site conditions: short growth season, high winds, low temperatures, low precipitation, and often on barren rocky substrates where few other plants survive. As a result the generalization ‘adversity begets longevity’ has had wide currency. But the next oldest species, giant sequoia (Sequoiadendron giganteum) grows at lower elevations then bristlecone, in dense forest on warmer, moister sites, among very tall specimens of several associated species (Harvey et al., 1980). Other long-lived trees can be found under either set of conditions, so Schulman’s generalization is unpersuasive. While bristlecone’s longevity is largely due to escape from fire, and those of its predators that cannot tolerate harsh mountain climates, giant sequoia owes its longevity to pest-repellent secondary defensive compounds and the fire protection afforded by thick bark. Thus, the two most famous of oldsters, whose natural ranges are separated by just a few miles, follow divergent strategies for living a long time. But both are trees, and what they have in common is much greater than what sets them apart. A tree manages to live for many years by predictably repeating a pattern of behavior, year after year, that has been refined by natural selection. Listed below are some of those key behaviors that make long life possible.
2. Creating a structure that can be expanded First a tree must create a structure that can progressively be expanded, so it can hold itself rigidly erect above other vegetation, and bathe in sunlight. The following discussion describes briefly the major growth activities of a tree (Wilson, 1970; Kozlowski, 1971). The growth of bark from the cork cambium will not be considered here. It should be borne in mind that rates and timing of growth are controlled mainly by environmental factors, but the growth patterns described below are genetically determined. 2.1. Growth in height Trees get taller by adding new growth increments to the ends of their vertical shoots, or ‘leaders’. This is referred to as primary growth. The new growth originates from the division and enlargement of cells localized in apical meristems. In the temperate zones, most trees’ apical meristems are located at the summits of dormant buds and are protected by a cover of bud scales, until the bud axis expands in the spring. Tropical trees’ apical meristems are usually protected by a looser rosette of embryonic leaves. The dome-shaped apical meristem is the locus of formation for leaf primordia. As these primordia enlarge, growth just below the bud axis—in the subapical region—forces the apex upwards in a surge of growth that is at first slow, then rapid, and then slow until it ceases. Maturation of the new tissues is wavelike, beginning in the lower (therefore older) part of the new shoot, proceeding upwards into the upper (therefore younger) part. The apical meristem typically ends the growth cycle by forming a new bud for the next cycle. Formation of a bud never ‘uses up’ the meristematic tissue of the apical meristem, but in some species a floral structure may form, thus precluding further lengthening of that axis. The final act of maturation is the lignification of the shoot’s supportive tissues, making the shoot ‘woody’.
656
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
This is accomplished by the deposition of lignin inside the cellulose walls of their cells (Panshin and de Zeeuw, 1970). In the early stage of shoot elongation it is typical for the new leaves to receive photosynthate from storage tissues in the previous year’s growth of the tree, but as the season of growth progresses, the maturing leaves become more self-sufficient, and finally produce a surplus of carbohydrate which is translocated to new storage tissues. This is usually in the ray parenchyma or axial parenchyma of the new woody shoots. In conifers, the needle leaves also serve as storage organs (Kozlowski, 1971). Meanwhile, other shoots are growing outwards in the same way, increasing the spread and complexity of the tree’s crown. 2.2. Growth in diameter As a tree grows taller, its trunk must thicken, or it will collapse. Diameter growth of trunks and limbs is by the division and enlargement of cells of a lateral meristem, the vascular cambium, a process called secondary growth. The vascular cambium, or cambium, as it is usually known, is a multicell-thick sheath or jacket of undifferentiated meristematic cells found between the outermost annual ring and the inner bark. When activated, the cambium produces daughter cells to its inside that rapidly differentiate into xylem (wood) cells. In broadleaved trees (angiosperms), most of these become fibers, which have a support function; or vessel cells, which align to form the pipelines that conduct xylem sap. In conifers, which have simpler wood with fewer cell types, both support and xylem sap transport are combined in tracheids. These are long, narrow cells tapered at both ends and aligned longitudinally in the stem. Vessel cells and tracheids formed early in the season (earlywood) have thinner walls and larger diameters than those formed later (latewood). The contrast between latewood of 1 year and earlywood of the next year delineates the ‘annual ring’ in cross-sections of temperate-zone trees. Tropical trees have indistinct non-annual rings that reflect episodes of growth. Daughter cells that form at the cambium’s outer periphery differentiate as phloem, or inner-bark cells. Most of them become sieve cells or sieve tube members which specialize in the polar transport of photosynthate, growth substances, etc. formed in the tree’s leaves and distributed throughout the tree. Unlike xylem sap-conducting tracheids and vessels, which must be dead and emptied of their contents in order to function, sieve cells must be alive to function. Much less phloem than xylem is formed in a growth cycle. Phloem cells typically live only 1 year, and are crushed against the wall of bark by outward expansion of the xylem. New phloem is formed at the beginning of the next growth cycle. Though conductive xylem cells die in their first year, ray parenchyma and axial parenchyma associated with them live for several years, even decades (Connor and Lanner, 1990). As new growth rings are added, older ones back in the stem experience death of their ray parenchyma, deposition of phenolics, etc., and frequently, discoloration (Panshin and de Zeeuw, 1970). The outer wood is referred to as sapwood, the inner as heartwood. Heartwood is usually considered to be biologically inert. 2.3. Root growth The primary growth of roots differs from that of stems by not being from buds, but from exposed apical meristems. They usually grow mostly when shoot growth is quiescent.
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
657
Secondary growth resembles that of stems, though many fine roots die shortly after forming and never become woody. Symbiotic relationships with mycorrhiza-forming fungi or with nitrogen-fixing bacteria or actinomycetes, increase the efficiency and probably extend the lives of many tree roots. The long-term goal of a tree is to become large enough and live long enough to produce seed. In order to do that it follows a dual strategy of first capturing space by rapidly growing upwards and outwards in prematurity; and then exploiting that space by filling it with foliage-bearing branches (Edelin, 1977). Analogous territorial claims continue underground, through a ‘heterorhizic’ system of long roots (or pioneer roots) which invade new ground, and dense mats or fans of short roots, which exploit the new ground between them (Sutton, 1969).
3. Making the structure redundant A structure that has more components than it needs at one time is regarded here as having the property of redundancy. There are several ways in which a tree may neutralize damage or loss of vital parts in order to live another day; or even perpetuate its genotype by replicating it. 3.1. Epicormic branching and stump sprouting as a response to injury Trees form many buds that fail to elongate into shoots the next year. Instead, they remain dormant. Buds are formed in the angle where a leaf is joined to a stem, so dormant buds are at risk of being overwhelmed by the radial growth of the stem, and embedded within it. However, the vascular trace that supplies sap to the bud grows a little each year, and allows the bud to remain alive while staying at the outer periphery of each year’s new sapwood. The buds’ growth is held in check by inhibitory growth substances produced in the leaves and moved downwards in the phloem. If the crown is damaged in a way that interrupts the flow of the growth substances, the dormant buds are released, and form new shoots that emerge from the stem surface. Termed epicormic branches, these shoots are a common response to pruning, wind breakage, defoliation, sudden exposure to bright sunlight, and other mechanical injuries to tree crowns. Epicormic branches are virtually ubiquitous in broadleaved trees, and have been considered relatively uncommon in conifers. During its sojourn beneath the bark, the dormant bud (or apical meristem, if the bud is sufficiently undeveloped) may branch profusely, forming a large number of potential shoots. This can produce a mass of new shoots very rapidly upon injury. In a few pines, buds long held dormant beneath the bark of the trunk sprout after fire scorches the tree’s crown. These include Canary Islands pine (P. canariensis), pond pine (P. serotina) (Fig. 2), pitch pine (P. rigida), and Chihuahua pine (P. leiophylla). A type of epicormic branch unique to pines is the interfoliar shoot that develops from the apical meristems within the needle bundle, or bundles, just proximal to the site of injury. Cooperrider (1938) described in detail the occurrence of this phenomenon in ponderosa pines (P. ponderosa) in the southwestern US that had been browsed by cattle. It is also a common response to defoliation injury.
658
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
Fig. 2. Pond pine is one of several pines that produce epicormic branches on their trunks following fire damage. Dismal Swamp, Virginia.
Another example of epicormics specialized to the needs of a particular tree species is the sequestering of innumerable buds on the upper surface of branches of the California endemic bigcone-spruce (Pseudotsuga macrocarpa). This species grows in highly combustible vegetation communities. When fire scorch kills its primary branches, they are soon clothed by a dense growth of newly sprouted foliage. This habit has been noted even in trees up to 300 years of age (Lanner, 1999) (Fig. 3). If an entire tree is badly injured, shoots of epicormic origin may emerge from the base of the trunk. Or if a tree is cut down, such shoots (called ‘stump sprouts’) may emerge from the cut stump. This is the basis of an ancient forestry practice, called ‘coppicing’, which consists of periodically felling trees for small wood products, and encouraging the growth of new coppice shoots for another crop. If the branches are cut off, instead of the whole tree, new branches sprout just below the point of shearing, a practice termed ‘pollarding’. Pollarding has been used as a means of growing firewood. Even when coppicing is not intended, it can occur on a large scale, as in the northern hardwood forests of the northeastern US, where heavy cutting of oaks, maples, and
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
659
Fig. 3. Bigcone spruce produces great numbers of epicormic branches following scorching by brush fires. Los Padres National Forest, California.
other trees that sprout vigorously, produced an extensive second-growth forest in the 19th century. A variant of the stump sprouts so common in broadleaved trees, but less so in conifers, is the formation of ‘fairy rings’ of fast-growing redwood (Sequoia sempervirens) sprouts surrounding the base of a dead tree, or the stump of one that has been cut (Barbour et al., 2001). It can hardly be doubted that epicormic branches prolong the life of the trees that produce them following injury; and that stump sprouting prolongs the life of its genotype. The question of prolongation of life via cloning will be considered below. 3.2. Epicormic branching as an ontogenetic shift While epicormic branching in broadleaved trees is usually a response to injury, it has been shown in recent years that some conifers form epicormics in the absence of trauma.
660
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
Fig. 4. A stand of old-growth western larch in which original primary branches have largely been replaced by epicormic branches. See especially the three trees left of center. Glacier National Park, Montana.
The primary branches of western larch (Larix occidentalis)—those formed on the trunk— become a liability that can limit tree age. As a branch ages it gets longer, more branched, and heavier. It is additionally burdened by the weight of rainwater, lichen growth, snow, bird nests, witches’ brooms, and wind torque. It can act like a sail in the wind, increasing susceptibility to windthrow, if it is well-anchored; or breaking off and leaving a gaping wound if not. But as such a branch ages, clusters of new green epicormics sprout from its base. In a typical 300–500-year-old western larch, two thirds of the crown may consist of long, dead primary limbs reaching out starkly, while masses of epicormics hug the trunk (Fig. 4). In the upper crown, the younger, smaller, more productive primary branches still prevail; but for how much longer is not known. Eventually, all the tree’s foliage may be borne on epicormics. This shift in the ontogeny of larch branches—what I refer to as an ontogenetic shift—allows the larch to maintain a crown of vigorous branches, surely extending its life for many years (Lanner, 1996). Several years before I reported this habit in western larch, Bryan and Lanner (1981) described it in Rocky Mountain Douglas-fir and coast Douglas-fir (Pseudotsuga menziesii vars. glauca and menziesii). We pointed out that there was apparently a ‘genotypically based program of crown maintenance that materially prolongs the life of the tree, and that this program requires no exogenous triggering mechanism’. Both in the Douglas-firs and the western larches, there were no indications of trauma having stimulated the epicormics, as is the case in broadleaved trees. Further study of old-growth coast Douglas-fir (Ishii and Ford, 2001) has confirmed that ‘epicormic shoot production maintains shoots and foliage of old P. menziesii trees after height growth and crown expansion have stopped, and may contribute to prolonging tree longevity’.
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
661
I have observed this phenomenon also in European larch (L. decidua), subalpine larch (L. lyallii), giant sequoia (Sequoiadendron giganteum), coast redwood, Sitka spruce (Picea sitchensis), and western hemlock (Tsuga heterophylla). There is in the pines an analogous situation to this, in which interfoliar branches are formed from within needle bundles without being stimulated by trauma. This occurs routinely in the bristlecone and foxtail pines of Pinus subsection Balfourianae (Connor and Lanner, 1987). 3.3. Sectoriality of the vascular system A tree’s woody skeleton can be viewed as a bundle of pipes gathered in the middle into a trunk, and radiating at either end into the root and branch systems (Shinozaki et al., 1964). One can further envision that the transport of water from the soil into a specified branch takes place only or mainly or largely in the trunk pipe or pipes that connect that branch to ‘its’ root pipe or pipes. In other words, there is a minimum of tangential spreading from one flow-way into another. If that were the case, the tree would be a composite of several hydraulically independent sectors. There is evidence that in some trees sectoriality of the vascular system does occur. For example, Rudinsky and Vité (1959) injected dyes into the sapstream of several conifer species. By examining the occurrence of the dye in different parts of the stem cross-section at different heights, they inferred five patterns of water ascent. Among these were ‘sectorial, winding ascent’, and ‘sectorial, straight ascent’. The spirality of ascent was correlated with the alignment of xylem cells. More recently, Larson et al. (1994) have shown that in very old northern white-cedars (Thuja occidentalis) that have been dye-injected, pathways of water transport appear to be sectored. Sectoriality of the vascular system is also indicated by field observations of the so-called ‘strip-bark’ habit of Great Basin bristlecone pine (Lanner, 1984, 1999) and ‘partial cambial mortality’ of Rocky Mountain bristlecone pine (P. aristata) (Schauer et al., 2001). Among very old trees one commonly observes trunks of bare wood that still retain only a narrow strip of bark. Beneath the bark, a live cambium produces annually a narrow strip of xylem and phloem. These strips can often be seen to extend downwards onto a major exposed root, and to wind around it into the soil. The bark strip ascends the trunk onto a branch containing some of the meager foliage still remaining on the tree (Fig. 5). The death of the major roots that no longer support live wood and bark has been attributed to drying following exposure through soil erosion, in the case of Great Basin bristlecone pine (Lanner, 1984, 1999), and perhaps to wind action in the case of Rocky Mountain bristlecone pine (Schauer et al., 2001). In the context of this discussion, it would appear that sectoriality of the vascular system of these trees is sufficient to prolong a tree’s old age by allowing it to keep alive ever-diminishing portions of its structure. 3.4. Root system replacement Under unusual circumstances, whole root systems of trees can be replaced. A well-known example is in coast redwood (Barbour et al., 2001). Trees growing on flood plains are
662
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
Fig. 5. A 1500-year-old Great Basin bristlecone pine, mostly dead due to exposure of its roots by soil erosion, with one surviving root providing water to a few branches via a narrow surviving strip of living wood and bark. Dixie National Forest, Utah.
susceptible to anaerobic soil conditions when their root systems are buried in deep sediments. Temporary relief is brought by upwards growth of new root tips. Permanent adjustment to these conditions comes about when an entirely new system of lateral roots forms within the sediment layer. Repeated floods over centuries can result in several such tiers of roots. The writer has observed an analogous situation in Hawaii Volcanoes National Park, where ohia lehua trees (Metrosideros polymorpha) partially buried in volcanic ash have formed new root systems in the upper ash layer which replace the buried root system. In both these cases, it can be safely inferred that the new roots aided tree longevity. 3.5. Clone formation Many trees can propagate vegetatively, forming small to large clones whose individual ‘ramets’ can substantially outlive the original mother tree. Griggs (1938) has described
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
663
how lower limbs of subalpine fir (Abies lasiocarpa) can ‘layer’, that is, form roots when in contact with moist soil, sometimes forming a circle of rooted branches around the parent tree. These layers can turn their tips upward, growing like saplings. Eventually, the branch connection with the parent tree is severed by decay, and the layers are physiologically independent. When the parent tree finally dies, the much younger layers prolong the life of the genotype, forming hollow-centered ‘timber atolls’ around the site of the dead parent tree. This also occurs in high-elevation Engelmann spruce (Picea engelmannii) (Wardle, 1968). An interesting variation of layering occurs in northern Quebec, with black spruce (P. mariana). Payette and Delwaide (1994) have described layered individuals at the extreme of the species’ range of occurrence that have persisted clonally during a period unfavorable to sexual reproduction, yet are still producing cones. A variant of layering occurs when a fallen tree retains roots in the soil and stays alive, while some of its branches grow upwards like trunks; and the trunk forms roots beneath the branch bases. Eventually, the fallen tree dies and decays, but its rooted branches persist as individuals. This has been described in Nepal alder (Alnus nepalensis) on the Island of Hawaii (Lanner, 1964). Perhaps the ultimate extreme of redundancy provided through vegetative propagation is the formation of large clones by root-sprouting. Nowhere is this more obvious than in quaking aspen (Populus tremuloides) in the Rocky Mountains (Lanner, 1984) (Fig. 6). Clones are formed by the sprouting of stems from the root systems of aspens that at some time in the past became established from seed. Over the years, these groves of genetically identical trees spread, as each ramet grows its own roots, and produces further ramets. Clones of thousands of interconnected stems can occupy 100 acres or more, and slowly migrate
Fig. 6. A clonal group of quaking aspen trees that have arisen as root sprouts from an interconnected root system. Note the similarity in branching angle and bark color of the stems. Wasatch-Cache National Forest, Utah.
664
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
over the landscape, as they grow at one edge while perhaps dying out at another. Minor morphological or phenological differences between clones allow easy identification of each clone’s boundaries. Throughout the Rocky Mountains it is difficult to find seedling aspens, as regeneration from root sprouts is highly predominant. It is widely believed that aspen has survived in the Rockies since the ice age only because of its clonal habit (Cottam, 1954), because present conditions seldom allow for seedling survival. However, after the catastrophic forest fires of 1988 in Yellowstone National Park, seedlings were reported to be abundant. Many other trees and shrubs form root-sprout clones, but seldom on the scale of quaking aspen. If genotypes whose seedling requirements are temporarily unmet in a changing environment can form clones, they may be able to survive until conditions permit the resumption of sexual reproduction. Thus, in effect, cloning is a prolongation of the longevity of the genotype.
Fig. 7. The trunk of a tree develops as a ‘beam of uniform resistance’, thinner at the top and thicker at the base to resist stresses without waste of material. Wind-flexing stimulates cambial activity, thus, open-grown trees like this ponderosa pine have thicker, more tapered trunks than trees growing in dense stands. Dixie National Forest, Utah.
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
665
4. ‘Optimizing’ the structure Trees must obey physical laws. They are subject to constant gravitational force, and are impacted by wind, snow loads, soil creep, earthquakes, and other instabilities. Fortunately, they have evolved means of defeating some of the forces that threaten to topple them. Mattheck (1991) has referred to these means as adaptive growth that optimizes the structure by evening load distribution. A familiar example of load optimization is the taper of long conifer trunks, due to more wood being grown in the lower trunk where stress is greatest, and the least wood at the treetop. In this way, the trunk, a beam of uniform resistance (Büsgen and Münch, 1929), is as strong as it needs to be, and as economical in the use of its nutrients as it can be (Fig. 7). Mattheck (1991) places under the rubric of minimizing external loading such diverse tree growth responses as the bowing upwards of a leaning trunk, the upright growth of a limb below the point where a leading shoot has broken off (Fig. 8), apical dominance of leading shoots over lateral shoots in order to prevent the latter from encroaching on the former, and
Fig. 8. A limb of this limber pine swept upwards many years ago in response to the death of the tree’s leader. Humboldt National Forest, Nevada.
666
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
Fig. 9. Recumbent stems of Mugo pine are swept upwards by the pushing force of compression wood on their lower surfaces. Swiss Alps.
growth of stems directed towards a light source. Wounds that create ‘notch stresses’ are viewed as having their healing regulated by the mechanical stresses manifested as ‘force fields’. The cases of leaning stems and upsweeping branches are examples of corrections in tree or branch orientation by means of ‘reaction wood’. The variant of reaction wood known as compression wood forms on the underside of conifer stems and exerts a pushing force that tends to raise or uphold the stem (Figs. 9 and 10). Compression wood cells are distinctive, and form dark, crescent-shaped peripheries of the growth rings in which they occur. The other variant, ‘tension wood’, forms on the upper side of leaning angiosperm stems and branches, and is also anatomically distinct. It exerts a pulling force that tends to raise or uphold the stem in which it forms (Kozlowski, 1971). Mattheck (1991) views much of adaptive growth as an effort to rearrange the centre of gravity of a disturbed tree above the rootstock. He points out that much of this adaptation is accomplished by the variability in how new wood increments are distributed on the tree.
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
667
Fig. 10. A blue spruce that nearly fell into the Snake River when the bank collapsed two years earlier has already started to turn its tip upwards, due to the force exerted by compression wood, and phototropic growth of the new shoots. Grand Teton National Park, Wyoming.
Mattheck’s views comprise a modern mechanical engineering perspective on the mechanical design of trees, and deserve careful attention by biologists.
5. Hormonal coordination of growth activities It is obvious that the tree depicted here is not a mere stick of inanimate wood stuck into the ground. Rather, it is an exquisitely fine-tuned and reactive organism that contains within itself the means to regulate its own shape and size. For many years it has been felt that the plant growth hormone auxin, or indole-3-acetic acid (IAA) is responsible for a wide range of regulatory processes in trees, as well as in other plants. Experimental research has succeeded in identifying numerous such effects, but others have proven complex and difficult to demonstrate in a definitive fashion. For example, apical dominance effects have been largely attributed to the action of auxin, but experimental results
668
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
Fig. 11. The loss of apical control by the upper parts of a fallen but still rooted Fremont cottonwood has resulted in the vigorous vertical growth of its lateral branches. Green River, Utah.
have implicated the possible involvement of gibberelins, ethylene, and cytokinins as well (Wilson, 2000). Auxin moves in polar fashion down the tree from sites in the foliage where most of it is produced (Uggla, 1998). So it is not unexpected that apical control of shoots by shoots higher in the crown has been shown due largely to auxin (Wilson, 2000). And the obverse of this relationship, the release of inhibited buds when crown damage interrupts the polar transport of auxin, is a common explanation for epicormics and stump sprouts. This same effect is also the classical explanation for the release of root sprouts that results in widespread clones like those of quaking aspen (Lanner, 1984), and the upright growth of lateral branches on fallen trees (Fig. 11). Auxin has also been implicated in the formation of compression wood in conifers (Timell, 1986). Several studies have shown increased IAA concentrations on the lower sides of tilted stems, correlating with the development of compression wood at those sites; or that application of the auxin-antagonist TIBA prevents the formation of compression wood (Funada et al., 1990). Perhaps the most important role of auxin in trees is its function as a ‘long distance positional signal’ emanating from the tree’s crown and being received throughout the stem (Uggla et al., 1996; Uggla, 1998). In this process, auxin synthesized mainly in the new developing leaves moves in polar fashion down the phloem and affects the growth of cambial cells. It does not directly regulate rates of cell division, but rather the number of cells that are dividing. In this way it coordinates cambial growth with apical growth, an essential function for creating a tree of balanced form. A related need, which appears also to be served by auxin at least in part, is the hormonedirected transport of structural carbohydrates from storage areas into the new, growing shoots of the tree. The more new leaves and stem internodes between them that are unfolding,
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
669
the longer the new shoot will be; and the greater its need for structural material. It appears that auxin synthesized in the leaves moves downwards, and directs the upward movement of stored nutrients. Lanner and Connor (1988) showed that in ponderosa pine, elongation of the shoot was regulated by the needle bundles, presumed to be the source of auxin, rather than by the shoot apical meristem. Auxin may also participate in the control of photosynthetic rates due to strong sink activity (Tschaplinski and Blake, 1989; Wardlaw, 1990). It seems a reasonable conjecture that the rapid feedbacks in increased photosynthetic rates correlated with high sink demand (Dickson, 1991); or the decreases in photosynthetic rates following reduction of sink demand (Myers et al., 1999) are auxin-mediated at least in part. Though much research remains to be done to further elucidate the complex role of plant hormones in the regulation of tree growth and form, it is already apparent that growth regulators, and particularly IAA, are of primary importance.
6. Defending the structure from enemies This topic was thoroughly and creatively reviewed a number of years ago by Loehle (1988), so I will summarize some of the major points brought out in his study. Loehle analyzed energy partitioning between defensive investments and growth in trees, to test the hypothesis that greater longevity should require an increased energy investment in such defence measures as defensive chemicals and thick bark. He assumed there is a trade-off between using energy for rapid growth, and for defensive compounds. His data were from published information on 159 American tree species, both broadleaved and coniferous. Longevity of broadleaved trees was correlated with their investment in defences, but that of conifers was not. Longevity of conifers was, however, correlated with resistance to wood decay. Since his comparisons were cost-benefit analyses between tree species with differing life spans, they are not directly applicable to the question asked here. However, they do bring attention to the importance of defensive chemicals synthesized in trees to defend against threats to structural integrity posed by insect and fungal pests. Broadleaved trees of greater longevity spent a larger proportion of their lives in the reproductive phase than did short-lived species. Among conifers there was no such relationship, though there was some indication that in very long-lived conifers the attainment of sexual maturity might be contingent upon attaining a minimum size rather than a minimum age.
7. Conclusions Trees are long-lived organisms. It is advantageous for them to be so, because sustained presence in an environment gives them many more opportunities to produce and disperse propagules. Thus, it is assumed here that selection for relative longevity has the effect of increasing fitness. There are many behaviors and structural characteristics of trees that enable them to live longer than any other class of organisms. Those discussed here include annual modular growth from persistent meristems of stem-cell-like parenchyma cells, resulting in a new tree being created each year atop the tissues of the old one; redundancy of branch,
670
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
trunk, and root outgrowth either in the presence or absence of trauma, resulting in large-scale replacement of dead or worn-out organs; a sectored vascular system that postpones death; formation of clones whose ramets are individually short-lived, but in the aggregate are potentially immortal; adaptive growth that optimizes the mechanical structure of the tree, allowing it to regain verticality, use material sparingly, and resist mechanical forces; with all of these behaviors controlled and coordinated, it appears, by hormonal means; and the synthesis of defensive compounds to defeat insect and fungal enemies. Not all trees possess all the above behaviors, though some are common to all. And some life-prolonging characteristics have been omitted from this discussion due to time and space constraints. These include, but are not limited to, the compartmentalization of decay in trees, whereby rots are confined to tissues that pre-date the wound opening the entry court; the healing over of injuries by periderms and callus tissue; and the incorporation through natural grafting, of the root systems of suppressed trees into those of dominants in the stand. Since all these diverse characteristics found among trees make their own contributions to longevity, it would seem fruitless to try to ascribe tree longevity to a single gene or gene complex. What makes a tree long-lived is simply being a tree. References Barbour, M., Lydon, S., Borchert, M., Popper, M., Whitworth, B., Evarts, J., 2001. Coast Redwood, A Natural and Cultural History. Cachuma Press, Los Olivos. Bryan, J.A., Lanner, R.M., 1981. Epicormic branching in Rocky Mountain Douglas-fir. Can. J. For. Res. 11, 190–199. Büsgen, M., Münch, E., 1929. The Structure And Life of Forest Trees, 3rd Edition. Chapman & Hall, London. Connor, K.F., Lanner, R.M., 1987. The architectural significance of interfoliar branches in Pinus subsection Balfourianae. Can. J. For. Res. 17, 269–272. Connor, K.F., Lanner, R.M., 1990. Effects of tree age on secondary xylem and phloem anatomy in stems of Great Basin bristlecone pine (Pinus longaeva). Am. J. Bot. 77, 1070–1077. Cooperrider, C.K., 1938. Recovery processes of ponderosa pine reproduction following injury to young annual growth. Plant. Phys. 13, 5–27. Cottam, W., 1954. Prevernal leafing of aspen in Utah mountains. J. Arnold Arbor. 35, 239–248. Currey, D.R., 1965. An ancient bristlecone pine stand in eastern Nevada. Ecology 46, 564–566. Dickson, R.E., 1991. Assimilate distribution and storage. In: Raghavendra, A.S. (Ed.), Physiology of Trees. Wiley, New York, pp. 51–85. Edelin, C., 1977. Images de l’Architecture des Conifères. Thesis, Univ. des Sci. et Tech. Du Languedoc, Montpelier. Funada, R., Mizukami, E., Kubo, T., Fushitani, M., Sugiyama, T., 1990. Distribution of indole-3-acetic acid and compression wood formation in the stems of inclined Cryptomeria japonica. Holzforschung 44, 331–334. Griggs, R.F., 1938. Timberlines in the northern Rocky Mountains. Ecology 19, 548–564. Harvey, H.T., Shellhammer, H.S., Stecker, R.E., 1980. Giant Sequoia Ecology, Fire and Reproduction. Sci. Monogr. Ser. No. 12, National Park Service, Washington, D.C. Ishii, H., Ford, E.D., 2001. The role of epicormic shoot production in maintaining foliage in old Pseudotsuga menziesii (Douglas-fir) trees. Can. J. Bot. 79, 251–264. Kozlowski, T.T., 1971. Growth and Development of Trees, Vols. 1/2. Academic Press, New York. Lanner, R.M., 1964. Clones of Nepal alder in Hawaii. J. Forestry 62, 636–637. Lanner, R.M., 1984. Trees of the Great Basin. Nevada University Press, Reno. Lanner, R.M., 1996.The role of epicormic branches in the life history of western larch. In: Schmidt, W.C., McDonald, K.J., (Eds.), Proceedings of an International Symposium on Ecology and Management of Larix Forests: A Look Ahead, 5–9 October 1992, Whitefish, Montana. USDA Forest Service Intermountain Research Station Gen. Tech. Rep. 319, 323–326.
R.M. Lanner / Ageing Research Reviews 1 (2002) 653–671
671
Lanner, R.M., 1999. Conifers of California. Cachuma Press, Los Olivos. Lanner, R.M., Connor, K.F., 1988. Control of shoot elongation in ponderosa pine: relative roles of apical and axillary meristems. Tree Physiol. 4, 233–243. Lanner, R.M., Connor, K.F., 2001. Does bristlecone pine senesce? Exp. Gerontol. 36, 675–685. Larson, D.W., Doubt, J., Matthes-Sears, U., 1994. Radially sectored hydraulic pathways in the xylem of Thuja occidentalis as revealed by the use of dyes. Intl. J. Plant Sci. 155, 569–582. Loehle, C., 1988. Tree life history strategies: the role of defences. Can. J. For. Res. 18, 209–222. Mattheck, C., 1991. Trees, the Mechanical Design. Springer, Berlin. Panshin, A.J., de Zeeuw, C., 1970. Textbook of Wood Technology, 3rd Edition, Vol. 1. McGraw-Hill, New York. Payette, S., Delwaide, A., 1994. Growth of black spruce at its northern limit in arctic Quebec, Canada. Arc. Alp. Res. 26, 174–179. Rudinsky, J.A., Vité, J.P., 1959. Certain ecological and phylogenetic aspects of the pattern of water circulation in conifers. Forest Sci. 5, 259–266. Schauer, A.J., Schoettle, A.W., Boyce, R.L., 2001. Partial cambial mortality in high-elevation Pinus aristata (Pinaceae). Am. J. Bot. 88, 646–652. Schopmeyer, C.S., 1974. Seeds of Woody Plants in the US. Agriculture Handbook no. 450, USDA Forest Service, Washington, D.C. Schulman, E., 1954. Longevity under adversity in conifers. Science 119, 396–399. Schulman, E., 1958. Bristlecone pine, oldest known living thing. Natl. Geogr. Mag. 113, 355–372. Shinozaki, K., Yoda, K., Hozumi, K., Kira, T., 1964. A quantitative analysis of plant form—the pipe model theory. I. Basic analyses. Jpn. J. Ecol. 14, 97–105. Sutton, R.F., 1969. Form and Development of Root Systems. Tech. Comm. No. 7, Commonwealth Forestry Bureau, Commonwealth Agric. Bur., Farnham Royal. Timell, T.E., 1986. Compression Wood in Gymnosperms. Springer, Berlin. Tschaplinski, T.J., Blake, T.J., 1989. The role of sink demand in carbon partitioning and photosynthetic reinvigoration following shoot decapitation. Physiologia Plant 75, 166–173. Uggla, C., 1998. New Perspectives on the Role of Auxin in Wood Formation. Acta Univ. Agric. Sueciae, Silvestria 58, Umeå. Uggla, C., Moritz, T., Sandberg, G., Sundberg, B., 1996. Auxin as a positional signal in pattern formation in plants. Proc. Natl. Acad. Sci. U.S.A. 93, 9282–9286. Wardlaw, I.F., 1990. The control of carbon partitioning in plants. New Phytol. 116, 341–381. Wardle, P., 1968. Engelmann spruce (Picea engelmannii Engelm.) at its upper limits on the Front Range, Colorado. Ecology 49, 483–495. Wilson, B.F., 1970. The Growing Tree. Massachusetts University Press, Amherst. Wilson, B.F., 2000. Apical control of branch growth. Am. J. Bot. 87, 601–607.