After tweaking some aspects of my study design, I returned to my study area 3 times over the past two weeks for some more data collection. I recorded data from 10 new replicates (based on the “rule of 10”) from the SE flank of Rainbow Mountain, as well as an additional 10 from the same are but at a lower elevation. I will not be including the data I collected a month ago for my “initial data collection assignment” in my final report (there are several reasons for this, including the fact that initially I was collecting my measurements too close to the ground, between 1 & 1.5m).
I have begun collecting my data from between 4-5m to mitigate possible confounding factors, including the slope angle near the base of the trees. This proved challenging at first, as making measurements higher up the tree was initially difficult to do with any degree of accuracy. I brought along a stepladder and a tape measure to assist with my measurements this time out, and after some practice I was able to devise a system for counting branches higher up the trees. I also used a different app to collect sunlight data to record in different units (watts per meter squared), which I think will provide a better representation of my predictor variable (“sunlight received”).
At the suggestion of professor Hebert, I also began taking measurements of the distance to the nearest neighbouring trees, as their presence may be a confounding variable in the growth of branches on the replicates being studied. In selecting the “nearest neighbour” I deemed only those trees that were 5m or taller to qualify, as any trees smaller than this would be unlikely to block sunlight from potentially reaching the replicates.
During my “initial observations” assignment, I was collecting on a day with some clouds, and their passing between taking measurements would create large inconsistencies in my light readings, even within the two sides of the same tree. In order to ensure the most uniform measurements of light, I collected on days with similar weather (clear, no clouds), and at the same time of day (12:00). (My first day of data collection took place at 14:00, so I returned at a later date to repeat the light measurements).
Field note book measurements
I noticed several nuances during my data collection that complicated the process more than I initially anticipated: The first one being that trees don’t always grow perfectly vertical. They often grow at an angle, which can make placement of the light meter somewhat difficult. Secondly, the nature of light filtering through a forest means that a slight difference in where the meter is placed can have vast implications on the reading it generates (i.e. the difference of being directly in a sunbeam or in the shade can be a matter of only a few cm). And furthermore, the location of where light filters through changes constantly throughout the day. Being consistent with the location of the light meter and time of data collection, as well as trying to move quickly without allowing haste to affect the quality was all I could do to ensure uniformity of results.
The topography in the lower elevation study area varied somewhat from the upper one, as did the species that populated it. While the slope was fairly uniform at 850m, closer to the valley bottom at 610m there existed many rolling microfeatures (small knolls) that affected the ways the trees caught the light. I chose to continue with my randomized sampling method in both areas, however it was more difficult to come across the species I was studying (Pseudotsuga menzeisii) at the lower elevation area, and several times I would have to re-enter compass bearing and number of paces in order to find a replicate. This was not an issue at the higher elevation.
One ancillary pattern I noticed during my data collection was that it is not merely the number of branches that seems asymmetrical on the two sides of the trees, but also the length and foliage of branches as well. While the data collection seems to have strengthened my belief in the prediction that more branches grow on the downhill side of the trees, they also seem to be significantly longer as well as more likely to be covered in foliage. I did not notice this trend until well into my data collection, and did not take any measurements regarding branch length however, as it would be quite difficult to do at a height of 4-5 meters and I was unprepared to do so. It is an interesting pattern nevertheless and I will consider if there is a way I can return to incorporate it into the project going forward.
Example of a replicate with asymmetrical foliage
