I collected my initial data on September 23, 2019; most of the collecting went smoothly and as planned, but during the collection and upon reflection some of it requires modification.
One of these difficulties was my sampling strategy. I had planned on using a systematic sampling strategy (explained in my Small Assignment 1) where I would use a random number generator to indicate which initial tree I would sample (at the very top and bottom row of trees on the hill). From there, I would count 9 trees, and 18 trees to the left and right of the initial sample to collect a total of 5 replicates. I realized while I was there; however, that if the randomly generated number was very small or very large, there may not be enough trees on one side of that tree to collect the second and third sample. I decided I would instead count another 9 (or 18, if necessary) trees after the 18th tree on the other side, to make up for this. I am having difficulties determining if this modified method of systematic sampling is “random” enough, yet I can’t think of an alternative.
With regards to the actual data I collected, the mean soil moisture levels at the bottom and top of the hill followed my prediction (the bottom of the hill having a higher mean moisture level). My leaf class strategy (class 1 being trees with 0-5% yellow leaves, class 6 being trees with 95-100% yellow leaves) will likely become more useful as the Fall progresses, as the vast majority of trees at the top and bottom of the hill all fell under the class 1 category. As well, the soil pH readings were very similar at both locations along the elevation gradient, not providing any useful measurements at the moment.
None of this is surprising, as the data I am collecting is likely to change drastically as the Fall progresses. For instance, we are now in our second day of snowfall in Calgary, Alberta, so it should be interesting to study the differences in soil moisture, pH, and leaf colour later this week after a few days of melting. For this reason, I plan to continue the same measurements as in my initial collection to allow for potential changes in pH, soil moisture, and leaf colour to be captured.
However, there are a few modifications to my data collection I would like to make. Firstly, I would like to adapt my data collection to the seasonal progression changes. For instance, as the Fall progresses I will likely add in a leaf loss measurement similar to my leaf class strategy. This way, I can measure the rate of leaf loss on top and bottom hill trees, which will be more applicable than leaf colour at that point. I may add other measurements as well (such as snow depth, to measure water infiltration rate). I also do not want to stick to a specific schedule as to when I collect data. Of course, data collection must be frequent enough to be able to capture changes in the health parameters I’ve mentioned, but I would also like to respond to weather changes. For instance, I will not be collecting data while the location is buried in 1ft of snow, as my pH and moisture meter do not have the capabilities to function in such drastic conditions. Instead I will collect data soon after the snow has melted, to study initial differences among the elevation gradient. Ideally, I would like to collect data no more frequently than every 5 days, but no less frequently than every 10 days (if weather permits).
I believe these modifications will improve my research because they will account for the natural and uncontrollable fluctuations in weather. Modifying the data I will be collecting will allow for my data to stay relevant as the Fall season progresses. Modifying the frequency of data collection will achieve the exact same thing. I hope, through these modifications across time, that I will more holistically be able to capture any differences in tree health among trees located at the bottom and top of Whispering Woods hill.
That’s all for now!
Madeleine Browne
