Blog Post 6: Data Collection
Data collection: My approach to data collection has been systematic as random selection does not apply well to the setting of my project and haphazard sampling yields less significant results in a statistical sense. I am measuring the growth of a thermophilic, photosynthetic, sulfur-utilizing bacteria Chloroflexus aurantiacus which I discovered growing near a wastewater output in a forest outside of Canmore Alberta. I want to determine if the quantity and density of bacterial growth are related to water temperature and depth and if so, what kind of relationship it holds (inverse, exponential, logarithmic, etc.). I started by using a 10cm x 10cm quadrat and holding it 5cm above the water and taking a photo directly above the quadrat. Within the quadrat, I would measure the water temperature, water depth, approximate water flow speed, and the reservoir width at that location along the bank. I can then use thresholding via ImageJ to determine the density of bacterial growth within the quadrat. I used systematic measurements, spacing samples every 5m along the reservoir banks. The reservoir became partially frozen so I jumped to the moving portion on the other side of the reservoir and continued sampling in 5m increments until the stream joined a much larger river (Bow river).
Replicates: In total, 10 locations were sampled and each location was sampled in triplicate. This leads to a total of 30 samples taken. The triplicate samples were taken adjacent on either side and in the middle of the 5m marker to represent the same area but different bacterial densities.
Problems: The largest problem has been the frozen section in the middle of the reservoir. Because the water slows down and collects, it freezes easily leaving a large gap in the middle of the sampled locations. However, the sample areas leading up to the ice have bacterial densities below 5% and thus it is suspected there is little to no growth in that area in the winter. Additionally, holding the quadrat in each photo, sometimes the tips of my fingers will be inside the quadrat which obscures about 1cm2 which does affect the density reading. It is of course omitted from the analysis but that does leave a tiny unit of area missing from the overall assessment. This impacts the overall reading by ~1% which is not critical but it is an area of improvement.
Ancillary Patterns: Before taking down data, I noticed that the amount of bacterial growth seemed to be related to the width of the reservoir and to the water flow. Because I noticed this potential association, I took down rough data with each sample though it was not a part of my initial hypothesis. My initial hypothesis still remains the primary focus of this field project but it will be interesting to analyze these additional factors to investigate whether a relationship exists.