Dendrochronology: Investigating the recent past
The red text is prepared by Ms. Chris Marion. © C. Marion 2005
Dendron (tree) + Chronos (time) = Dendrochronology: The use of tree rings as time markers: Dendrochronology can be a very powerful tool in answering ecological questions about the recent past. It also has applications in archaeology, geomorphology, forestry, climatology, and even law. This document is only an introduction to this technique; refer to Henri Grissino-Mayer's Ultimate Tree-Ring Web Pages (http://web.utk.edu/~grissino/) or to F. Schweingruber's 1988 Tree Rings: Basics and Applications of Dendrochronology for a much more comprehensive coverage of the technique and its many applications.
1) Introduction to wood anatomy and tree ring formation Wood, or secondary xylem, is laid down in trees during the growing season to the inside of the cambium layer, accounting for lateral growth of trees. A tree ring is composed of two (more or less distinct) bands of cells (figure 1). The earlywood, the light-coloured band, is laid down in the spring and early summer, when water availability is highest. The xylem cells produced by the cambium are then rather large in diameter, and have thin walls. Latewood is produced later on in the summer and in the early fall. Latewood cells are somewhat smaller than the early wood cells, and have a much thicker cell wall and much smaller lumen, accounting for the darker color of the latewood. At the end of the growing season, wood production shuts down until the following spring, when large, thin-walled cells are again produced by the cambium, making a very sharp contrast with the previous year's dark, tight latewood (at least in conifers!) (Figure 2).
Water availability and warmth of the growing season are the two main factors affecting tree ring width. A wet, warm season will lead to the formation of wide, light-coloured bands in most boreal evergreen trees, with ring width generally increasing with the length of the favourable growing season. Dry or cold summers will result in narrower rings. Although the trees' response to growing conditions is species-specific and also depends on other factors such as nutrient and light availability, the previous generalizations usually apply to most of our boreal conifer species. Deciduous species are more difficult to work with as the latewood is not much different form the early wood, and observation of the cells themselves is sometimes the only way to distinguish the annual rings.
Many events will happen in the life of a tree which may be recorded in its wood either at the cellular level or as more obvious scars or deformations in the tree itself. Frost, insect epidemics, and droughts are a few examples of events that will be recorded in the rings at the precise year (or series of years) at which they occurred. Fire and other cambium-destroying accidents will stop the production of wood where the cambium was damaged, leaving a scar that might eventually close with time; this scar can also be dated. Cutting down a tree will stop lateral growth altogether and tell us the year the tree was felled. Tree death by flooding, burying, or uprooting can also be dated.
Sampling, preparation, and dating of tree cores
Sampling This can be done in a few ways. It is possible to obtain cores from trees using an increment borer. This type of sampling does not kill the tree, but can be difficult if the trees are rotten. Careful crossdating of the samples collected this way is necessary, because very narrow or incomplete/missing rings may be missed since the cores represent only a small fraction of the trees' cross-section. Taking a tree disk (or 'cookie') destroys the tree but gives the best sample as incomplete rings can be detected, narrow rings can sometimes be seen better somewhere else around the stem, and patches of rot can be avoided when dating the disk. Events such as fire scars are almost impossible to date without a disk.
Preparation Wood samples from increment borer are glued into core holders, then dried and carefully sanded so that the annual rings may be readily observed. (The late and early wood is difficult to distinguish in some trees, such as trembling Aspen. These may be dyed to bring out the different grain pattern.)
Counting and dating Dating live trees is very straightforward: the last year laid down is the current year's wood production. Counting back from the outside ring towards the inside of the disk will give an approximate age for the tree. (In forestry practice, trees are very rarely sampled at the root collar although many rings may be missing from the beginning of a tree's life with a sample taken at chest height. Sampling as close to the ground as possible helps avoid underestimating the age of a tree.) Counting from the inside out towards a scar will then allow dating of the scar. Dead trees can be cross-dated with live trees to allow for age determination. Use on small dot on the core holder by each decade, two dots by every fifty years and three dots by each century. This process helps in keeping track of the number of rings counted. Pencil marks are recommended over pinpricks as they can be erased should a dating mistake be detected.
The first step in any research project or study that relys on dendrochronology for answers is to decide exactly what the question is we want to answer. This is important as it will influence the species we will sample, the sites where we will sample them, the number of samples we need, and the kind of samples we take. For example, a study wanting to investigate past outbreaks of the spruce budworm will involve sampling trees sensitive to defoliation by the insect (fir and white spruce) as well as trees that are not sensitive to it (pine) to distinguish the insect signal from any background climate signal. A project looking for links between tree-ring width and recorded rainfall to make predictions of past precipitation regimes will involve sampling trees that are very sensitive to variation in rainfall: species that need a lot of water, or trees growing on sites where water from precipitation is a limiting factor, will give the best indication of precipitation levels.