User: | Open Learning Faculty Member:
My study will be examining the density of coniferous trees in proximity to an adjacent creek (specifically Pinus ponderosa). The environmental gradient was chosen from the riparian bank to the upland forest apex, with three locations observed along the gradient; the riparian area, flood plain, and upland forest.
Observations were recorded in the field notes below, and in essence describe the density of tree species in relation to the area. It was observed that broad leaf deciduous trees dominate in the riparian area, with a noticeable lack of coniferous growth. The complete inverse was noted in the upland forest, with only Ponderosa pine inhabiting and a more diverse mixture of species in the adjoining flood plain.
I since formulated the hypotheses that the density of Pinus ponderosa decreases with proximity to Mission Creek within the observation area. A potential response variable would be tree concentration, while a predictor variable may be distance from riparian bank, or possibly the attributed soil conditions. These natural variables are both continuous, and would therefore justify a regression analysis.
Sounds like you have a good start to your study! I have had students measure soil moisture in various ways and measure it at each plot vegetation is measured. An alternative is to keep your predicter variable as distance and to soil moisture qualitatively along your gradient and report those results separately while still allowing you to speculate about the role of soil moisture in your discussion.