Dendrochronology Lesson #1: Shrub growth rings aren’t obvious

I can’t say I wasn’t warned.

Tree dendrochronology was pioneered in the early 1930’s by A.E. Douglass, who then founded the Laboratory for Tree-Ring Research at the University of Arizona. Since that time, tree rings have been used in archaeology, chemistry, ecology, and biogeography to reconstruct the timing of past events. Growth rings in trees are, for many species, clearly detectable – sometimes even with the naked eye! – and tree dendrochronology techniques are well established. This is not the case with shrubs, whose structure and growth form tends to complicate the researcher’s ability to accurately age an individual. Shrub dendrochronology is still in its pioneering days.

The challenge with shrubs is that although they typically add a layer of new cells each growing season (i.e., summer), this doesn’t consistently result in a new layer of wood on each stem. Being multi-stemmed, shrubs allocate resources where they are needed. As a result, some stems grow more than others within a season, including sometimes not putting on any growth at all that year. Some stems may die, while new shoot spring forth; a stem may fall to the ground only to reproduce clonally or that stem may become covered in soil and find more purpose in life as a root. The end result is that growth rings in shrubs are quite often very narrow, partial, entirely missing, or lacking the distinction between the lighter (larger-celled) wood of early-season growth and the darker (denser) wood of late-season growth. As commented by Myers-Smith and Hik (2017. Journal of Ecology, 106: 547-560), two leading researchers in the field of shrub dendrochronology: “Willow stem sections varied in the ease at which rings could be counted and the information that these rings contained.

It was thus with trepidation that I began preparing my 1,760 stem samples for growth rate analysis. Determining the growth rate of a stem is pretty straightforward: Count the number of rings at the base to find out how many years ago that stem emerged from the ground, then count the number of rings halfway up the longest stem and use the difference in rings (years) to get an estimate of how long it took that stem to grow to half its length; relate this back to the total length of the stem to estimate growth rate. All I needed to do was to identify and count growth rings.

Optimistically, I began sanding stems in late December, beginning with 120 grit sandpaper and polishing off the stem at 320 grit. It was encouraging to notice what appeared to be growth rings! With luck, I could pop these stems under a magnifying glass, easily count the rings, and be on my way to analyzing growth rate. But that wasn’t the case. Even under the lab’s high-powered magnifying glass, clear growth rings weren’t evident. Counting rings was going to take some work.



Next, I soaked a pair of stem samples in water overnight and the following day I sliced off very thin sections of 30 µm, 25 µm, and 20 µm from each sample. For reference, 10 µm (microns, or micrometers) is one one-hundredth of a millimeter! Using a dissecting microscope, I could discern rings – but barely. Hovering over the dissecting microscope, fingers playing with the fine-focus adjustment, and my eyes watering, I attempted to count rings: One, two, three? Four? Or was that still three? Four, five, six… {blink} Which one was six again? Start over: One, two, OK three and four are the same thing, so next is four, five, six, seven, eight? Is that late-season growth or rot? Nine? Darn! {start over}. And that was the process for samples with more obvious rings!

Staining sections with less obvious rings in dark soya sauce helped make the rings slightly more discernable… but only sometimes. Photographing the microscope view, playing with the contrast in the resulting image, and counting rings from the computer screen also helped… a little, and not always. For each stem section, I aimed to count growth rings across two radii (each radius being at least 90o from the other) on each of two slices of differing thickness (choosing the best among 30 µm, 25 µm, or 20 µm). Looking back at my data was discouraging: while some ring counts only differed by a ring or two, my estimates were incredibly inaccurate for many stems. To count 36, 39, 37, and 32 rings on a single sample – in other words, for that sample to vary by 7 years or 20% of its lifetime – isn’t acceptable. Counting rings is going to take a lot of work!

Upon meeting with my PhD supervisor, Dr. Ryan Danby, he cheerfully sketched out a projected graph of my growth ring learning curve, helpfully pointing out my current position in the process relative to the point at which the heavens will open and light will shine down, clearly and accurately illuminating shrub growth rings: 

With an optimistic mind I thus experiment with different equipment and techniques for counting shrub growth rings in search of the combination that works best for me. One, two, three… four, five? Yes, five; six, seven…

One thought on “Dendrochronology Lesson #1: Shrub growth rings aren’t obvious

Leave a Reply

Your email address will not be published. Required fields are marked *