Gymnosperm Database: information on tree age determination

Tree age determination can be a fairly complicated process. There are basically two methods that can be used: (1) tree rings, and (2) everything else. We'll look at (2) first because it's a relatively simple problem.

Everything Else

The best way of determining a tree's age is to find out when it was planted. Obviously, this is difficult, but it occasionally works if the tree was planted by humans in a context where historical information can provide us with a date. For example, a forestry plantation, a tree in an extensive garden, or an exotic tree planted when an area was first colonized, could all be aged from historical data. For some kinds of trees, such as cycads, palm trees, yuccas or giant cacti, historical information may provide the only means of getting a realistic age.

Dating a tree without rings can also be done by measuring its growth rate or by using a chemical analysis, such as radiocarbon dating. Growth rate measurements tend to assume that the growth rate measured over a given recent time period can be extrapolated to the entire lifespan of the plant. As we will see when discussing tree rings, such estimates can be wildly inaccurate. Nonetheless, it has been used to estimate ages of 1000 to 4000 (!) years for quite a few yew trees in England (Hartzell 1991) and it was used by Chamberlain (1919) to infer an age of 2000 years for a cycad based on counting the number of leaf scars on the trunk and multiplying by an estimate of how long it took the plant to produce a new leaf. A similar method has been applied to estimate the age of palm trees. Similarly, a widely quoted age estimate of 200 years for a typical saguaro cactus (Carnegia gigantea) is based on observing how much a relatively large saguaro can grow over a period of a few years and then extrapolating. In this database, ages obtained by extrapolation are cited only as historical curiosities and are assumed to be inaccurate.

Carbon dating has occasionally been used to measure tree ages. It is used surprisingly often by scientists who are unaware of the uses of dendrochronology, and is quite useful with certain trees native to the wet tropics, where there is a minimal amount of seasonal weather variation. In that climate, some trees just never go through the seasonal period of reduced or halted growth that causes a tree ring to form. Radiocarbon (Carbon-14) dates can be reliable if the carbon in the heart of the tree is stable -- that is, if it has remained in place since the tree started its growth. This seems to be a valid assumption for a hardwood tree with no heartrot and intact heartwood. However, some other trees -- such as palms -- do not have a single definable area of their trunk that dates to the tree's early years, and such trees cannot be carbon dated. Supposedly a radiocarbon age of 2000 years has been obtained for Welwitschia mirabilis, but I have been unable to find details on this.

Tree Rings

Nearly all tree ages cited in this database are derived from tree rings. The principle used here is that in most trees that form rings, the rings are formed annually, so the number of rings in the tree will provide a fairly close approximation of the tree's age. In practice, there are a number of problems with this principle: (1) trees occasionally produce more than one ring a year; (2) trees occasionally go a year or more without producing a ring; (3) you have to somehow see the rings to count them, preferably without killing the tree; (4) there are a host of problems dealing with how the tree was first established and how fast it grew in its earliest years; and (5) if these problems force you to accept some uncertainties in the age estimate, then how can you estimate those uncertainties and perhaps compensate for them? We'll look at these problems in turn, but first, a little information on how a tree grows.

The Life of a Tree

Plants are not born in the same way that most animals are. At the very beginning, a tree might be born either from a seed, or from a growing part of another tree. If it grows from a seed, then at least it is possible to say that in year n there was no tree, and in year n+1 there was a tree; it had a definite beginning, and in theory we could determine when it was. If, on the other hand, the tree arose from a growing part of another tree by the process of vegetative (or clonal) reproduction, then things are more complicated. The tree may have a arisen from a branch, or from a root, that was some number of years old and that was part of another tree. It's not a small problem. Vegetative reproduction is common in gymnosperms, and in some groups it is much more common than reproduction from seed. For example, most of the giant California redwoods (Sequoia sempervirens) probably originated from roots of their forefathers, as do most cycads of the genus Encephalartos. In such cases, continuity of living plant material may exceed the life of observed individuals by thousands of years.

In any event, at some point we can say there is a young plant growing, though its age may already be unclear. Let's say it's a young tree and it will be producing annual rings. As it gets bigger, it produces more and more foliage. That foliage requires more and more water and that requires wider and wider rings to carry the water from the roots to the foliage. If you look at the stump of a young tree, you will see this process written in its rings -- the rings are very narrow when the tree is small, but they get wider and wider with each successive year. If the tree is growing in the open and has sufficient light and water, this process will continue for decades, carrying the tree through seedling and sapling stages until its a fine tall tree (incidentally, this phenomenon of rapid youthful growth provides the economic basis for commercial tree farming).

At some point, the tree will start to grow more slowly. It may be getting close to its maximum height, or it may be encountering competition from other trees. Whatever the reason, eventually it stops putting on more and more foliage and reaches a relatively steady state. Every year, it carries about as much foliage and uses about as much water as the year before. Once the tree reaches this stage, each annual ring that it produces will have about the same cross-sectional area as the previous annual ring. However, because of the width of the ring, that area will be spread out around a tree that is a little bit larger. Consequently, each ring is a tiny bit narrower than the ring before it. For most big conifers, this process can go on for hundreds of years. This explains why it is foolish to extrapolate a tree's age on the basis of relatively recent growth -- most fairly old trees are putting on narrower rings now than at any previous time in their lives. [someday I'll put in a digression here on how some trees beat the system by changing their stem geometry]

The problems with age estimation using tree rings

(1) Trees occasionally produce more than one ring a year. This has to do with how a tree ring is formed. Most conifer tree rings are light-colored on the inside and dark-colored on the outside; this alternation of light and dark is what makes the ring easy to see. The change in color occurs because early in the growing season, the tree produces large cells; as the growing season goes on, drought stress causes the tree to produce smaller cells. Because the cells are smaller, there is proportionally more cell wall material, and this causes the cells to appear darker. If there is a period of renewed rainfall in the later part of the growing season, the tree may start to produce big cells again, and then small cells a bit later on as drought stress resumes. The effect is to produce a second ring, commonly called a false ring. The most striking example of this that I have heard of involved Pinus caribaea growing in (I think) the Dominican Republic under a climate that had very little seasonal variation. These pines would put on a ring every time there was a wet spell, and over a period of about 30 years they averaged 4 to 5 rings a year.

Fortunately, in temperate and subtropical climates it is usually possible to spot false rings by detailed microscopic examination of the cell structure of the tree ring. It's hard to describe exactly what you have to look for; suffice it to say that after you've seen it 1,000 (or 10,000, or more) times, you get a pretty good idea of what to look for. This is what graduate students in dendrochronology do to earn their daily gruel.

(2) Trees occasionally go a year or more without producing a ring. It happens because the tree suffers some sort of severe stress. For example, the tree could be struck by lightning, burned by a fire, attacked by insects, injured by human activity, or under stress due to adverse weather (such as extreme cold or a severe drought). It can be very difficult to detect missing rings. It's done by means of a procedure called crossdating, which I will not explain because a great link exists to do just that:

CROSSDATING

By crossdating, you can use trees that don't have missing rings to find where other trees DO have missing rings. Ah, the alert reader asks,"what if all of the trees are missing the ring for a year?" The answer is, that doesn't quite happen -- but it can come close. When they were first putting together a long bristlecone pine (Pinus longaeva) chronology there was a year, I think it was about 605 AD, that was missing -- and it took hundreds of samples before that year finally turned up. It was a very dry, very cold year, but there were a few trees growing in sheltered locations that managed to form a ring anyway. Incidentally, some trees live in very harsh situations and have a lot of missing rings. The general feeling among dendrochronologists seems to be that if more than 10 percent of the rings are missing, it is very difficult to figure out the dating of a specimen and consequently there is little confidence in age estimates from such specimens. Such age estimates can probably be assumed to underestimate the tree's true age.

(3) You have to somehow see the rings to count them, preferably without killing the tree. There are two common approaches to this problem. One is a saw. The oldest tree ever known was a bristlecone pine that was cut down with a chainsaw to determine how old it was. It was found to contain 4,995 rings. The remains of this little tree now reside at the Laboratory of Tree-Ring Research, where they continue to inspire people to not cut trees down just to find out how old they are. However, saws are very useful for sampling trees that are already dead. For example, the oldest known examples of Pacific silver fir (Abies amabilis) are actually based on counts of tree rings from stumps in clearcuts. Of course, these trees are no longer alive -- but the age data tell us how old they can get, and there are enough old-growth silver fir out there that comparably old trees are probably still alive.

Increment coring

The second way of seeing tree rings is with a tool called an increment borer (photo courtesy the Ultimate Tree-Ring Web Pages). It's a hollow drill that takes out a core about 4 mm in diameter and up to 50 cm long. There is a fair bit of debate about how much this hurts the tree. Without going into great detail, it apparently doesn't do much harm to a large and healthy tree, but may kill small or sickly ones. The rule of thumb is, do not core a tree without a very good reason, and then communicate your findings in a suitable forum (such as scientific journals) so that someone else will not have to repeat the damage a few years hence.

Finally, I should mention that people have examined tree rings without directly harming the tree by using techniques such as nuclear magnetic resonance tomography. There was a paper on this in Jacoby & Hornbeck (1987). As far as I know, though, nothing ever came of it. Nearly all live-tree maximum ages reported in this database are based on increment core data.

(4) How was the tree was first established and how fast did it grow in its earliest years; and (5) if these problems force you to accept some uncertainties in the age estimate, then how can you estimate those uncertainties and perhaps compensate for them? [well, this will have to be continued at a later date]

[home] [topics]

This page is from the Gymnosperm Database
URL: http://www.geocities.com/RainForest/Canopy/2285/topics/oldest.htm
Edited by Christopher J. Earle
E-mail:earlecj@earthlink.net
Last modified on 21-Nov-1999