Category: Post 8: Tables and Graphs

Good afternoon Professor Elliot & Class,

It was a challenge to organize and summarize my data, but once I had figured out what the most important parts of my research were, it was easier for me to visualize them. My prediction was that there would be a greater abundance of large woody vegetation (e.g. trees) on the eastern, westerly facing side of Jack Creek.

Hypothesis: If the landscape has a higher elevation, and is westerly facing, then a greater abundance of large woody vegetation will be present.
Prediction: A greater amount of large woody vegetation will be present in western facing slopes with a higher degree of aspect and elevation.
Response Variable: Large woody vegetation abundance
Explanatory (predictor) Variable: Elevation and aspect

During my field data collection, I took counts of both trees and shrubs to create comparison. I took an average of the vegetation count to the number of data points to create an accurate representation of the data. Below are two graphs representing the average number of species in response to elevation and aspect.

The outcome of my field studies was slightly different than I expected, however, it reveals that further exploration is necessary as to why there was a greater amount of vegetation in the flat meadows 15 metres on the western side of the creek. Other factors might influence the data, such as disturbance, opens fields, and amount of sunlight.

The table I created was relatively easy to compile.  The greatest challenge was in choosing what not to include.

I had collected location data on my sample subject of invertebrates, which I thought to include in a table or figure in some manner.  Including this data would have added unnecessary complexity to the table or figure and the reader would be unlikely to gain any meaningful insights from knowing that a species was recorded 3m down a transect and 6cm to the East.

In graphical form, the data makes a visual statement that species were most often found in a non-vegetated location.  I had expected species to be found most often in vegetation, so the visual representation is particularly stark to me. Also with the invertebrate abundance being more or less equal between sites, it is a visual representation which I find interesting.

I chose to use a graph to display the data that I have collected. I used a line graph with two lines, one for grazing the 10 grazing replicates and one for the 10 non-grazing replicates, to show the average amounts of damage in percentage to each of the replicate areas.I chose to display it as an percentage of the overall area because I think it is easier for readers to understand and interpret. I think that the line graph I selected clearly displays the difference in damage between the two treatments and the overall trend. I did not have a difficult time organizing my data as I collected it in an excel spreadsheet so that it was quick and easy to enter equations and form graphs and tables from it. The data stayed with my predicted trend more or less, but I would be interested to explore the affect that the number of geese present grazing has on the percent of area damaged. When I first visited the site sometime ago there were only 2-3 geese grazing, but on the day I collected the 20 replicates there were 5. I’d assume that the increase in geese presence would also increase the % of area damaged.

To summarize my findings for the richness of healthy living moss found on the trunks of trees I had to use a number scale to help average out my finds to show any difference in the moss richness found on the three tree top coverage categories. Once it was decided that a tree with hardly no moss present would be given a value of zero, a tree with some healthy moss present was 0.5, and a tree with high moss richness was 1.0 it was easy to summarize the data collected.

As shown in table 1 the partial tree top exposer category showed the highest average of moss richness, this went against the prediction that moss richness would be highest at the shelter tree top group. Though it was shown that the exposed tree top group did have the lowest moss richness where it was about half the amount of the other two tree categories which supports the prediction that the exposed tree top group would have the lowest moss richness. As the data was being collected over the two weeks it was observed that the partial tree top exposer group had increased the moss richness on the tree trunk the most noticeably. Where it is thought the partial tree top exposer will have the fastest rate of increased moss richness on the tree trunks when entering the spring season in British Columbia when compared to the other two tree groups. However, more data collected over a longer time period during the late winter and spring season is needed to test this prediction.

Table 1: Total average of  healthy living moss found on the tree trunks of the three tree top exposer categories. Each tree top category contained ten replicates where date was collected four times at each site over the spanned of two weeks. A number value of 0.0, 0.5, and 1.0 was used to stand for no moss richness, some moss richness, and high moss richness found on each of the tree trunks respectively.

For my project I have decided to use graphs to represent a summary of my raw data. Because of my statistical tests used, I created a graph that shows the mean number of brush bushes per square meter at various elevations from the creek bottom. This graph shows the differences between the six elevations I compared.  When analyzing my data, I decided to remove the highest 6m counted because I only had ~6 samples. This made my comparisons a lot more representative of the un-paved study area. I used a bar graph, and feel like this is the best way to visually represent how the density of brush bushes changes as the height from the creek increases.

I will also include a table denoting the significant differences between elevation categories for the big sagebrush bushes. There were no significant differences in the rabbitbrush bushes across the different elevations, so I will just report that in my results section.

This data took me a long time to go through and figure out how best to analyze it. I chose an ANOVA over t-tests because they are a more powerful statistical test. I also compared the valley bush distribution to a Poisson distribution.

The table I created included all of the data I collected over the five days. It was a summary of the various bird species abundance at each of the study sites. By producing a graph that summarizes all of the data collected, it enables for trends and patterns to be clearly outlined to draw conclusions from. However, from this data I found it to be confusing how to incorporate it into a graph to obtain a more visual demonstration of the species abundance over the three different sites.

Nevertheless, I noticed that each study site had different bird species associated with it. More specifically I noticed that each study site had a dominant species that inhabited the area. This was the outcome I was expecting, however, that leads to further questions such as why particular bird species favour one site over another? What is different about each site that a spatial gradient is created?

As a secondary research project it would be interesting to monitor the bird species and their migration patterns. Since the seasons are in the midst of changing the species abundance at each of the sites must also be changing; therefore, it would be interesting to determine which site the bird species favour as their migration patterns change. In addition, it would be interesting to see the displacement of the bird species as new birds migrate into the area. I’m assuming that depending on the bird species the migration patterns are different; therefore, if a smaller bird favoured a particular site it would be interesting to observe what would happen if a larger bird slowly started to inhabit the area.