In a growing season, woody vegetation (trees and shrubs) adds a new layer of cells called annual growth rings. The size of a ring is influenced by environmental conditions in that year, such as moisture and light, and occasional events such as disease, insect outbreak, or fire. Dendrochronology is the scientific technique used to assess changes in the growth of woody vegetation in response to environmental change. Annual growth rings provide a window into past growth and an opportunity to retrospectively compare with past climate conditions and determine important drivers of change.
Satellite images show that productivity of many tundra ecosystems has increased over the past 30 years. Field studies have linked these trends to increasing growth of woody vegetation, primarily shrubs. However, the specifics of that growth are still poorly understood. For example, is this trend due to new individuals having become established, and/or are growth rates increasing? Quantifying variability in growth patterns will further our understanding of vegetation change across the Bathurst caribou range.
Most of the Bathurst range exists above treeline, where shrubs dominate the landscape. Dwarf birch (Betula glandulosa), various willow (Salix spp.) species, and alder (Alnus crispa) are low-stature and multi-stemmed. Using dendrochronology to study deciduous shrubs is challenging, and required us to amend sampling and processing techniques, and to develop a suitable monitoring protocol that could be applied to our project.
Part I – Tree Rings
During the first year of our project, we sampled trees in the winter range of the Bathurst caribou herd to establish a record of growth over the past several hundred years. This record provides historical context for current vegetation growth taking place over the entire range. Samples were collected at two different locations, one near the core/western edge of the Bathurst range (Snare Lake) and one near the eastern edge (Lutsel K’e) (Figure 1). Both of these sampling sites are located near the forest line, an area whose land cover is defined by an approximate ratio of 1,000 units of tree cover to 1 unit of tundra vegetation.
Figure 1: Sampling sites within the Bathurst caribou range. Shaded areas show the annual (tan) and calving (yellow) grounds, respectively. Green lines show tree line (dotted) and forest line (solid). Important locations for sampling are labelled and potential future sampling areas are identified with cross-hatch.
At Snare Lake, we sampled 32 white spruce trees and 30 black spruce trees on the western shore of the lake. The white spruce were located within an open esker environment (a ridge of sand and gravel deposited during glacial times) and had an average age of 233 years with the oldest individual being 381 years old. The average growth rate of the white spruce trees was 0.57 mm/year. The black spruce were densely populated in an adjacent lower area with lichen ground cover. The average age of the black spruce were 178 years and the oldest individual was 226 years old. The black spruce showed a lower growth rate of 0.34 mm/year.
At Lutsel K’e we sampled 39 white spruce trees at two areas of exposed rock outcrops located about 4 km from each other. Here the average age was 160 years and the oldest individual was 298 years old. The average growth rate of the sampled white spruce in this region was similar to the white spruce at Snare Lake (0.56 mm/year).
Dendrochronological sampling extracts two drinking-straw sized (4.3 mm) cores of wood from a living tree without harm. These cores are mounted to wooden frames and then finely sanded to reveal annual rings, which are then counted and measured under a microscope. The average ring width in each year was plotted for each combination of site and tree species, giving us two 250-year chronologies for white spruce (one at each sampling site) and a ~200-year chronology from the black spruce at Snare Lake (Figure 3).
The chronologies show two interesting trends related to shifts in growth patterns, suggesting that growing conditions are changing. First, white spruce and black spruce at Snare Lake grew at similar rates during the 19th century, but over the past one hundred years black spruce have been growing more rapidly than white spruce at this site. Whatever environmental condition(s) had previously limited the growth of black spruce at this site appear to have shifted. Second, white spruce growth from Lutsel K’e and Snare Lake become more similar over time, particularly since the 1920s. This increased similarity of growth between sites located hundreds of kilometers apart suggests broad-scale drivers such as temperature are having an increasing influence on vegetation growth.
The next steps for this part of the project involve more in-depth comparisons to climatic data for the period of 1900-2010. Comparisons to climate will hopefully contribute toward a reconstruction of the primary driver(s) of growth over the past 250 years. In the second year we will also be expanding the coverage of our tree-ring network by including additional chronologies from within the fall and/or winter range of the Bathurst herd. This information will be valuable for comparisons with shrub growth from areas further north and will contribute to our understanding of the magnitude and rate of recent vegetation change.
Part II – Shrubs
In the summer of 2017 we visited the GNWT Tundra Ecological Research Station (TERS) located approximately 50 km north of the treeline at Daring Lake. There, we collected test samples of the most common shrub species we are likely to encounter across the Bathurst range (Figure 2) and developed a suitable shrub sampling protocol for our project.
Figure 2: Pictures showing various aspects of the shrub field sampling procedure.
Our sampling protocol considered a variety of factors, including growth forms (individual shrubs vs. large shrub mats), sampling location (i.e., above the root vs. at ground level vs. above ground), living vs. dead stems, and shrub canopy (coverage of the upper layer of leaves and stems). It also balanced the quality of sampling with the length of time required for sampling to ensure we could adequately sample the necessary number of individuals in a given area. We also worked on a method for establishing transects (linear survey plots) that will provide a representative snapshot of vegetation over a given area.
Returning to our lab, we sliced samples into thin-sections of about 40 µm (or 0.00004 of a metre) in thickness to permit a highly accurate assessment of annual rings, used a microscope to photograph each section, and then digitally stitched images together for analysis.
Overall, we found that annual rings from the three main species of shrub could be used to clearly identify ring structures in samples with diameters greater than 5 mm, and collecting multiple sections per stem improved cross-dating between and within samples.
Next steps will include ring width analysis to identify years with very high or very low amounts of growth that may be indicative of years in which conditions were or were not beneficial for radial growth, and comparing annual growth to satellite measurements from the past two decades. Together this will provide a more complete picture of vegetation change across the Bathurst caribou range and provide a potential method for future environmental monitoring across the NWT.