Old Trees and the Meaning of ‘Old’

Abstract

Individual trees with a high age are a characteristic feature of old-growth forests. However, for plants, and here especially for trees, the terms senescence, age and even death are difficult to define. This is partly because humans tend to quantify tree age according to astronomical cycles, more specifically using annual tree rings, while for trees age is a relative quantity related to the pace of ontogenetic deve-lopment. The slower the biological processes unfold, the older individual trees become. This is why the oldest living trees with an age close to 5,000 years are found on extreme sites. Senescence is based on the principle of ‘programmed cell death’, the timing of which may vary between individual cells, tissues types, organs and whole plants. For example, an earlywood tracheid dies after 20 days before it becomes functional as water-conducting tissue, while a neighboring parenchyma cell may survive for 50 years before serving the purpose of mechanical stabilisation as a dead cell. The individual ramets (shoots) of a clonal tree die after 50 years, while the genet may survive for over 10,000 years. In this chapter, we critically discuss different concepts of longevity and review what is known about the variation in and limits to the longevity of woody plants.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

While the mere presence of ‘old’ trees does not automatically indicate old-growth conditions (see Chap. 2 by Wirth et al., this volume), it is fair to say that many old-growth forests contain a high number of trees close to their maximum longevity. Besides definitional aspects, tree longevity per se is a key demographic parameter controlling successional dynamics of species replacement, stand structure and biogeochemical cycles (see Chap. 5 by Wirth et al., this volume). This chapter takes a dendroecological perspective on tree longevity. The first part will explore differences in longevities between different life forms and will ask to what extent trees differ from herbs and shrubs and among each other (Sect. 3.2). The second part will discuss the mechanisms underlying the death of cells, tissues and whole plants (Sect. 3.3). It will be shown that the concept of death is problematic in the context of clonal plants, and that the inevitable presence of external mortality agents may bias our perception of biological limits of longevity.

2 Longevity of Conifers and Angiosperms

“After an individual becomes established, it must persist” (Weiher et al. 1999). The question remains: for how long? Undoubtedly, the oldest living beings on our planet are trees. The oldest trees look back on an individual history of almost 5,000 years, whereas most herbaceous plants persist for only a few years and some annuals die in the course of weeks. Apparently, longevity is highly variable among plants.

Reconstructing the age of an old tree is far from trivial because ring formation can be suppressed in stress periods or rings may be doubled in interrupted growing periods. In such cases, age determination requires the dendrochronological technique of cross-dating. As shown in Fig. 3.1, this simple method allows the determination of felling dates of ancient woods as well as the age determination of living trees.

Fig. 3.1

Principle of dendrochronological cross-dating. The key to evaluating the calendar date of the last ring on a stem disk is the irregular distribution of extreme years, the so called pointer years (Schweingruber et al. 2006)

A selection of the maximum ages of some of the oldest trees (see Table 3.1) shows that the availability of data on tree longevity, determined by cross-dating, is not evenly distributed across the world. The list suggests that tree longevity itself is not strictly related to the climate. The hot spot of tree longevity is located in the mountain ranges of western North America, where many species reach an age of 2,000 years. In contrast, the Canadian boreal forest is characterised by remarkably short maximum longevities. Here, conifers rarely exceed an age of 400 years. The biogeochemical relevance of these differences in longevity is shown in the model study presented in Wirth et al. (see Chap. 5 by Wirth et al., this volume). However, low longevities are not a feature of boreal forests in general, as some larches in the Eurasian subalpine zones and the boreal taiga are over 1,000 years old. The Eurasian Stone pines (Pinus cembra and Pinus sibirica) can probably also reach that age, but relevant dendrochronological data are missing. Spruces, firs, and deciduous trees do not exceed a maximum lifespan of 500 years. In this context, it is interesting to note that the oldest artificial tree, a cross-dated tree ring sequence composed of different individuals of central European living and subfossile oaks and pines is 12,460 years old (Friedrich et al. 2004).

Table 3.1 Selection of maximum (extreme) tree ages. Sources: Old-list, Rocky Mountain Tree-Ring Research (http://www.rmtrr.org/oldlist.htm), and tree-ring data-bank (http://www.wsl.ch), Dendrochronological laboratories of P. Gassmann, Neuchatel, Switzerland, and H. Egger, Boll, Switzerland

Information on the maximum longevity of shrubs is very limited, but it seems that they are generally shorter-lived than trees (Schweingruber 1995) and dwarf-shrubs (see below). The oldest known shrubs grow in Siberia. Hantemirov et al. (2003) found an 840-year-old Juniperus sibirica. Dendrochronological analyses in a dry temperate Quercus pubescens forest in the Swiss Jura mountains revealed that the age of the root stocks of several shrub species capable of resprouting is usually much higher than the age of the shoots. For Cornus sanguinea the ages of the root stock and the shoots were 35 years and 5 years, respectively; for Ribes alpinum the relationships was 62 vs 10 years; and for Lonicera xylosteum 48 vs 12 years.

More is known about the maximum longevities of dwarf shrubs. According to Kihlman (1890), Callaghan (1973) and Schweingruber and Poschlod (2005), the oldest individuals may reach maximum ages of up to 200 years (Table 3.2). Even a small, delicate plant such as Dryas integrifolia has been found to live for at least 145 years. In general, individuals of dwarf shrubs older than 50 years are not rare in subalpine and sub-Arctic environments.

Table 3.2 Selection of maximum ages of dwarf shrubs according Kihlman (1890), Callaghan (1973) and Schweingruber and Poschlod (2005)

Within the group of herbs, the age of the whole plant can be determined only in species that form a taproot -- this being the only structure where all rings are preserved. In clonally growing rhizomatous plants, counting of annual rings in the rhizomes allows the age of currently present tissues to be determined, but not the age of the whole plant. The maximum ages of tap-rooted herbs are well known for western Europe (Schweingruber and Poschlod 2005). As for the dwarf shrubs, the herbaceous species with highest longevities grow in the subalpine and alpine zone. We found 50 annual rings in Trifolium alpinum, 43 in Draba aizoides, 40 in Minuartia sedoides and 32 in Eritrichium nanum. The maximum age of the majority of tap-rooted herbaceous plants in the lowlands is between 1 and 6 years.

3 What Limits the Life Span of a Tree?

Different aspects of ageing have been discussed in a number of reviews. A summary is given in Schweingruber and Poschlod (2005). Most studies to date focus on physiological aging processes and refer to parameters at the level of cells, tissues or organs, while processes relevant at the level of the whole plant are usually ignored (Thomas et al. 2003; Zentgraf et al. 2004; Schweingruber et al. 2006).

3.1 Programmed Cell Death

The process of secondary growth in trees involves the continuous formation and death of cells. Programmed cell death creates a diverse array of cell longevities. Taking the xylem as an example, tracheids and vessels formed very early in the growing season may live for only a few days, while the same cell types formed later may survive for months. In general, however, all water-conducting tissues die at the end of the growing season. Non-conducting fibres normally die after cell-wall thickening is finished. Their lifespan is short and rarely exceeds 1 year. In contrast, most parenchyma cells are longer-lived. Axial and vertical parenchyma cells in the sapwood may live for several years. The maximum age of living ray cells in Robinia pseudoacacia is 4--6 years and up to 130 years in Sequoiadendron giganteum.

Trees face the problem that they can grow taller only by progressively putting on new cell layers around the entire surface of the stem. Over the years, this leads to the accumulation of a massive body of woody tissue, which, if containing live, respiring parenchyma cells (usually around 7% and 16% of the sapwood volume in conifers and hardwoods, respectively; White et al. 2000) would inevitably drain the energy resources of the tree even under the most favourable growing conditions due to the fact that the surface of assimilating foliage increases more slowly with size than the wood volume. To overcome this problem, old parenchymatic cells die and excrete fungicidal phenolic substances (Fig. 3.2). This protects the interior dead woody tissues from microbial decomposition, which is important in maintaining the mechanical stability of the tree [Fig. 3.3; but see Thomas (2000) for trees without true heartwood]. Often, this chemical impregnation of the heartwood goes along with a discoloration allowing us to distinguish macroscopically the coloured heartwood from the pale “living” sapwood.

Fig. 3.2

Microscopic section through the heartwood of the dwarf shrub Eriogonum jamesii. Axial parenchyma cells contain dark substances, probably phenols

Fig. 3.3

Sapwood and heartwood in the xylem of a Robinia pseudoacaciastem. All cell types in the dark part (heartwood) of the stem are dead and contain phenolic, fungicide substances. Water transport and storage of assimilates occur in the light part (sapwood). Axial and vertical (ray) parenchyma cells are living

The design of a tree crown is largely the product of cladaptosis, the die-back of twigs and branches. The process of cladaptosis is crucial for a trees ability to forage for light. It enables the tree to abscise branches that run into a negative carbon balance due to self-shading and light competition with neighbours. Some species, such as oaks and poplars, show a weak and almost unlignified zone at the base of the twigs, which acts as a predetermined breaking point (Fig. 3.4). Other species actively form a barrier zone at the base of their twigs to cut the twigs off from the water supply. As a consequence they dry up and drop off after a few months or years.

Fig. 3.4

Branches with scars of dropped twigs on Quercus robur. Crown formation is based on the existence of this process of cladaptosis

3.2 Whole Plant Longevity -- Internal Versus External Factors

There is little literature about the endogenous processes controlling the longevity of whole plants (Ricklefs and Finch 1995) and, if discussed, the focus is either on genetic components or on the mere quantification of mortality rates as a demographic parameter.

For herbs (with taproots -- see above) the data allow us at least to distinguish between annual and perennial species (Schweingruber and Poschlod 2005). In addition, this latter study demonstrated that the life span of most herbs is definitely restricted to a few years, because the genetic potential excludes the possibility of reaching longevities in the order of decades (Fig. 3.5).

Fig. 3.5

Maximum ages of central European herbs and dwarf shrubs. Black columns Number of species with taproots roots (total of 603 species), grey columns species with rhizomes (total of 232 species); 63% of the species with taproots have a limited age between 1 and 6 years, and only of 8% of the plants have a lifespan that exceeds 20 years (Schweingruber and Poschlod 2005)

The genetic predisposition of whole plant death is difficult to evaluate in long-lived trees, because it would require long-term common garden experiments that would by far exceed human longevity. The collection of maximum tree ages given in Table 3.1 is rather arbitrary. Moreover, the available data probably underestimate maximum longevities. So-called “age hunters” tend to search for trees with particularly thick stems, but we know very well that size is an unreliable predictor for tree age. Quite on the contrary, maximum tree ages are much lower on sites with optimal environmental conditions. Dendrochronologists have often found the oldest trees on marginal sites, where trees survive close to their ecological limit, e.g. in swamps or on shallow soils near the timberline. Such a negative relationship between site quality and longevity can be found in both ‘annual’ herbs and perennial trees. The ‘annual’ Linum catharticum completes its life cycle in 1 year only at optimal sites, but needs 3 years in the subalpine zone. The giant tree Sequoiadendron giganteum may grow for more than 3,000 years without any sign of senescence in its natural habitat in the Rocky Mountains, with ring widths remaining on average below 1 mm for centuries. In contrast, the same tree species grown in European plantations in a wet oceanic climate on deep soils has an average ring width of about 1 cm, but becomes very susceptible to wind storms. Thus, mortality seems to be correlated with size rather than absolute age.

Determination of maximum longevity becomes impossible in trees that reproduce clonally, such as poplars, willows and hornbeam. In these species, new ramets continue to sprout long after the initial stumps has decayed away. Even where the founder module is still present in the population of ramets, molecular methods may be required to actually identify it. This is illustrated by two examples: in the Canadian boreal forest, black spruce (Picea mariana) spreads vegetatively by branch layering. A dendrochronological analysis revealed that a genet having regenerated from seeds after a forest fire may reach an age of at least 300 years (Legère and Payette 1981). However, molecular studies showed that a larger genet could even reach 1,800 years (Laberge et al. 2000). The oldest genet on earth is a polycormon of Lomatia tasmanica in Western Australia spread over 1.2 km². Charcoal buried next to fossilised leaves with the same genome as the contemporary trees was dated as being at least 43,600 years old (Lynch et al. 1998).

It remains an open question whether trees are in principal immortal or whether their genetic constitution limits their lifespan as is the case for herbaceous plants with taproots. The example of Lomatia tasmanicain fact suggests that clonal tree species are almost immortal. However, even for non-clonal trees we are unable to know for sure whether they would not live forever (or at least for much longer), if they were protected from disturbances and diseases. While we know very little about the endogenous controls of longevity, there are countless studies on how various external agents such as fire, wind, flooding, herbivory, pathogens, pollutants, etc. speed up senescence and reduce the lifespan of trees. In the following we can only briefly touch on this topic, and we do so only to emphasise that the influence of external mortality factors biases our view of tree longevity.

Based on the simple observation that ecological factors limit the existence of single trees, we have to accept the old idea that trees often die by exhaustion or starvation (Molisch 1938), e.g. due to a lack of light (Fig. 3.6) or energy (i.e. summer temperatures; Fig. 3.7) or a shortage of water (Bigler et al. 2006). This may lead to false conclusions about the longevity of species. For example, maximum longevities reported in the literature for the Eurasian Betula pendularange between 120 and 140 years (Nikolov and Helmisaari 1992). However, Schulze et al. (2005) recently found individual trees older than 300 years. One reason for the low literature estimates may be that birches, as typical pioneer trees, tend to be out-competed by tall-statured late-successional species already after about 100 years. Thus, the majority of birches dies early as a result of light starvation and not because they have reached their biological limit. Older individuals may simply have been overlooked. Another example was already mentioned above: old trees are very rare in the Canadian boreal forest. However, this is determined not only by the biological age limit of the tree species, but also by the circumstance that in the North American boreal forest lethal crown fires recur on average every 100 years (see Chap. 13 by Bergeron and Harper, this volume, and Wirth 2005). Toxic substances, for example, sulphur dioxide from anthropogenic pollution sources can kill trees, but we have also found that the reaction to poisonous agents depends on the species and may vary even between individuals. Trees at the borderline of the catastrophic sulphur contamination in the downwind area of Norilsk (Siberia) clearly show species-, individual-, and site-dependent mortality: larches (Larix sibirica) in all ecological situations were dead, whereas spruces (Picea obovata) and birches (Betula pendula) growing at the same ecological sites were either dead, or had reduced foliage or even looked healthy. Spruces in the most intensive contaminated regions survived as dwarfs in moist riverbeds between healthy looking sedges (Schweingruber and Voronin 1996, see Fig. 3.8).

Fig. 3.6

Starvation due to light shortage. Competitive beeches have suppressed the crowns of pines (Pinus sylvestris) and induced their death. The starving period is indicated by the narrow rings with small latewood and the enhanced frequency of resin ducts in the pre-lethal period

Fig. 3.7

Dying at the beginning of the Little Ice Age between 1430 and 1450 AD. Stands of larches (Larix sibirica) died at the timberline in the Polar Ural. The stumps have remained and have been dated dendrochronologically (Shiyatov 1992)

Fig. 3.8

Death due to anthropogenic pollution near a smelter in Central Siberia (Norilsk). Extremely high SO₂ content in the air leads to selective tree death. The most sensitive species are larches (Larix sibirica); spruces (Picea obovata) and birches (Betula pendula) are less sensitive. Within Siberian spruce there are also intra-specific differences in sensitivity: some individuals die, but some manage to survive high doses of toxic gases

Biological degradation caused by mammals, insects, nematodes and fungi affects different species in different ways (Thomas and Sadras 2001). A morphological expression of the different sensitivities towards herbivory of pathogen attack is the formation of barrier zones (Schweingruber 2001, see Fig. 3.9). Longitudinal barriers are, for example, weak in birch and ash, but very effective in beech and maple (Dujesiefken and Liese 1991).

Fig. 3.9

Formation of a barrier zone after mechanical wounding of the cambial zone. The zone below the wound was laterally compartmentalised by the formation of a toxic barrier zone. Fungal decay occurs only in the part below the wound, all other parts are protected by the barrier zone. Arctostaphylos uva-ursi. 40x

A few years ago there was a great hope that tree-ring curves would allow the prediction of individual lifespan. Indeed, there is strong evidence that the risk of mortality is negatively correlated with growth and that the shape of this relationship differs between trees with low and high shade tolerance (Kobe et al. 1995). However, it is too simple to assume that a reduced growth period in adult trees would indicate senescence. Very narrow ring sequences simply indicate a transient period of starvation and, as such, are a reversible feature of tree growth (Fritts 1976). Moreover, tree death may occur abruptly or gradually. Rapid death has often been observed in shade-intolerant species, whereas shade-tolerant species literally shrink to death on a branch-by-branch basis over decades to centuries. This variability in behaviour makes it impossible to use tree ring sequences to infer estimates of tree longevity or even to predict the expected duration until death.

In summary, the large range of longevities realised by trees makes it likely that a genetic predisposition in general determines longevity, but the real lifespan will always be modified by the environment. Thus, separating ‘nature and nurture’ in their effect on longevity will remain a difficult task.

4 Concluding Remarks

Within the plant system, and within the range of life forms, trees are very special. Thanks to their high longevity, trees may accumulate enormous amounts of biomass. The largest tree on earth, a Sequoiadendron giganteumcontains 1,470 m³ wood with a dry weight of 800 tons. One single tree contains approximately 400 tons carbon. These ‘biological monsters’ would not exist if they were not perfectly designed to resist extreme mechanical stress (Fig. 3.10).

Fig. 3.10

Mammoth trees (Sequoiadendron giganteum) represent tremendous carbon stocks and may live for 3,000 years

The potential age of physically existing trees exceeds that of all other life forms. Old trees tend to be perfectly adapted to specific sites. In Europe, many old larches and stone pines at the alpine timberline germinated at the beginning of the Little Ice Age in the thirteenth century. They have survived many stress periods and are now benefitting from the current warming period. Since these ‘living fossils’ maintain the potential to regenerate generatively and, in many cases, also vegetatively, these trees are an indispensable genetic resource. Old trees do not lose their capacity to respond to the environment. Variations in the size of their cells and in the width of their tree rings demonstrate that even millennium-old trees maintain their biological sensitivity and their potential to react to environmental stress and favourable periods. Expressions of this reaction potential are e.g. scars, callus formation (Fig. 3.11), reaction wood (Fig. 3.12) and growth variations such as abrupt growth changes and pointer years.

Fig. 3.11

Frost ring. The reaction to extreme low temperatures at the beginning of the growing season at the end of June 1601 in the Polar Ural was the formation of a frost ring. Larix sibirica (100x)

Fig. 3.12

The reaction to an extreme change in the position of a branch after being hit by a stone was the formation of a callus and compression wood. Pinus mugo. (20x)

Thanks to their longevity and their sustained sensitivity, old trees represent important archives of past climates. Dendrochronological techniques allow the reconstruction of annual climatic patterns and the occurrence of extreme weather events at both local and global level. In doing so, they provide a means of placing the contemporary man-made climate warming into a historical context (Fritts 1976; Schweingruber 1995; Fig. 3.11).

Old trees have always fascinated people. Gollwitzer (1984) has summarised the evidence for the human fascination with old trees, which goes back at least 3,000 years: old trees were the seats of the gods. They stood at the centre of world religions and embodied myths. People celebrated and mourned under the canopy of old trees (Fig. 3.13). During the period of enlightenment in the seventeenth century, people began to study trees scientifically. Today, we still have not solved the puzzle of why trees become as old as they are. Only one thing is certain: the circumstance that so many tree species have gone extinct tells us that even trees do not live forever -- as palaeontology shows us (Zimmermann 1959).

Fig. 3.13

People celebrating under the canopy of an old lime (Fischbach und Masius 1879)