User: | Open Learning Faculty Member:
The biological attribute I have chosen to study in my selected area is the distribution of tree species. My area is inbetween a bike path and a steep hill so I divided the area into 3 separate areas, the location beside the bike trail, the middle, and the location at the base of the hill.
I found that in each area, the dominant tree species was very different. The area beside the bike trail consisted of mostly evergreen trees like subalpine fir and engelmann spruce (see picture below).
In the middle area, the dominant speces was black cottonwood and aspen (see picture below).
In the area at the base of the hill, there was very little large trees and an abundant mix of small alder trees and willow plants were present (see picture below).
While walking through the gradient, I noticed that the amount of water flowing through the bases of the trees increased while getting closer to the base of the hill. I also know that evergreens prefer dry well drained soil and cottonwood and alder/willow will grow in wetter areas. Therefore, my hypothesis is that the distribution of the most abundant tree species in each area is dependant on the water content of the soil. My prediction is that the water content in the soil will be highest at the base of the hill and lowest beside the bike trail.
Based on my hypothesis, a potential response variable would be the dominant tree species of the area and one explanatory variable would be the water content of the soil. I will test this by taking soil samples throughout the gradient. This would be a natural experiment (because you cannot change the predictor variable) and this experiment would have a logistic regression design as the predictor variable is continuous (soil moisture) and the response variable is categorical (tree species). I would test the soil moisture by taking soil samples in each of the 3 areas using a distance-based sampling procedure. I would then measure the weight of the soil when wet, dry it out in the oven, then measure it when dry.
If you are saying that soil moisture predicts tree species composition it will be better to have data to back up your observations about the dominant tree in each of your three areas. Even though your observations seem pretty clear, you usually measure your response variable.