Post 8. Tables and Graphs
My data was summarized in simple bar graphs. My separate graphs compared plant density, number of individuals and plant height between growth on the nurse logs and ground. Numbers from 21 samples were averaged out. The differences were quite noticeable; I didn’t realize how much until I graphed it. It certainly emphasizes the advantage of using a visual graphic over looking at charts of numbers.
The results of my field data were easy to summarize and visually represent in tables and graphs. The bar graph I submitted summarizes bird abundance (number of individuals) observed at the three different sites along the urban gradient representing different levels of urbanization. I predicted that bird abundance would follow a gradient with the lowest number of individuals observed in the urbanized area (Site 1) and the highest number of individuals observed in the natural area (Site 3) . When I initially graphed this data I found that the highest abundance was in fact at the most urban site. However, further examination of the data indicated that this was due to the large portion of observations (roughly 2/3) in the urban area that consisted of rock doves. As a result, the graph I created displays the overall abundance along the urbanization gradient but highlights the proportion of rock doves at each site so that the underlying trend (when removing rock doves from the examination at all sites in which they were observed) becomes apparent, which confirms my prediction.
The data I collected was easy to summarize in a bar graph. The outcome was slightly unexpected. When the stats were run (ANOVA), I determined that my result was not significant. As my sample size (n=5 for both conditions) was small, the standard deviation turned out to be quite large. This likely was why my data was insignificant. A method that would likely decrease deviation would be to collect more samples. However, as this is not a study I plan on publishing, I don’t want to cause additional disturbance to the natural environment.
The graph I prepared was a visual representation of the raw data collected during my field study. The results were as expected with the fesuce grass species growing at a faster rate in shaded areas, compared to in full or partial sun areas. With the quantity of data collected it was hard to categorize the data. Averages of the growth in each area would have been more beneficial however the graph does give you a full representation. In collecting the data I would be curious to explore further, how other grass varieties are effected by sun exposure. I would also be curious to see if ambient temperature and rainfall also have an effect, in addition to sun exposure.
Upon collection of my data, summarizing the information obtained was not a challenging task. However, what I did find challenging was organizing the data that I obtained into a manner that could be easily read and understood by others. A 1m2 quadrat was used to measure percentage coverage, a 0.5m2 quadrat was used to measure the abundance of the species and a 0.25m2 quadrat was used to measure absence/presence. Each of the three quadrats was placed randomly five times at each site, and data was collected. Overall, my data did not reveal anything too surprising. Although the results obtained supported my hypothesis and predictions to some extent, upon analysis of the data collected, I do realize that other factors that impact each of the plant species directly play a more essential role in the distribution. As such, further exploration ideas consist of some factors such as, climate, soil condition, water exposure, invasive species, biological interactions, and etc. An example of my data regarding abundance, organized into a table is shown below:
Table 1: The table below depicts the abundance of each of the six species present at each of the three sites at Milliken District Park. The first site, Site A, is within the forest, the second site, Site B, is outside of the forest, just before a large pond, and the third site, Site C, is the area on the other side of the pond. The ACFOR scale was used to measure whether the species was Abundant, Common, Frequent, Occasional, or Rare. A species was considered abundant if it was present 10 or more times within the quadrat. It was considered common if it was present 7-9 times within the quadrat, it was considered frequent if it was present 5-6 times within the quadrat, it was considered occasional if it was present 3-4 times within the quadrat and it was considered rare if it was present 2 or fewer times within the quadrat.
There was no difficulty organizing my data, although I had to decide if it would be best to do a bar graph or a line graph. In the end I went with a bar graph as that was the most familiar graph for me. The outcome was actually not quite as I had expected. I had initially assumed there would be more fruit on the plum tree as opposed to the other two trees but that did not seem to be the case here. In considering water stress alone, one might think that the idea of drought affecting fruit growth negatively is utterly untrue. That is why I chose to do further research on why these results were not what I had predicted and came across a few interesting explanations. Below is a table of all the weeks I counted fruit. The dashes under the Cherry Tree mean that during the 6th-12th week, there was no fruit on the tree as it had finished its harvest point.
| Week | 1st | 2nd | 3rd | 4th | 5th | 6th | 7th | 8th | 9th | 10th | 11th | 12th |
| Cherry | 10 | 30 | 45 | 57 | 57 | – | – | – | – | – | – | – |
| Pear | 5 | 16 | 30 | 36 | 45 | 57 | 65 | 72 | 75 | 79 | 80 | 80 |
| Plum | 0 | 0 | 6 | 12 | 20 | 29 | 31 | 33 | 36 | 37 | 39 | 39 |
My data has been difficult to put into a graph and summarize. The outcome was not what I had expected or predicted. This is in regards to the aquatic insect numbers increasing with number of plants. Plotting number of plants against number of aquatic insects created a random scatter plot with no real trend line.
The one relationship I can see is plotted in the graph above, but difficult to see without some explanation. Transects 1-5 has the most consistent amount of plant growth, while 6-12 have lower plant growth that is more sporadic. Transects 1-5 have higher number of insect species versus transects 6-12 (the transects are located across a shoreline from every 4’ from 1-12). Transect 7 and 9 throw things off a bit, since they have high plant numbers but lower levels of insect species. My thought is that the transects next to them have fewer plants, so more insects may gravitate towards the area of denser plants, transects 1-5. It’s possible some of the species may only stay in areas of more plant density.
As I stated in my last blog post 6, breaking the pond into plant densities on a larger scale, and random sampling each of those areas, would be a better design along with identification of each species.
Figure 1. Species richness as a daily value per observation site. This value is defined as a number of distinct species recorded on each visit. McArthur island park A is a site on the western exit from the park, and B is on the eastern. Riverside park A is located on southern entrance and B is located at the northwest shore. McDonald Park contained a single observation site, therefore, no data representing B site is present.
Before using any statistical analyses, it was important to see whether collected data had variation, in order to be a viable representative. I chose the scatterplot to show the distribution of means representing number of individual species observed at one visit. The pattern that can be seen on this graph is that the sites located in the same park have similar values for species richness. But the values between parks differ. Because these parks are located in three different urban environments, effects of urbanization were further compared with distribution patterns. One paper that I read stated that bees usually stick to one site even if a better one present within short range. This means that sites A and B for both parks should be assessed as independent even though the data seems very similar.
Because this graph was not supposed to have an actual x value, excel was not able to create it. MiniTab16 was used instead as it allows to set x value while changing y value.
Above is my table of information obtained during my sampling period of 10 different randomized replicates. Increasing my replicates from 2 to 10 has really helped increase the amount of data that I have been able to obtain. This has helped increase the amount of African Elephants captured on the camera traps and has given me more information to utilize in my graph. The data obtained is exciting to see as there is definitely a trend as to what temperature the elephants drink at. The status quo is that most animals will drink at the hottest time of the day in order to cool themselves down and quench there thirsty that would be thought to be peaking at that particular time of day. However the evidence that I have collected shows that a majority of the Elephants are drinking at 20 degrees C. A great majority of the more sporadic drinking habits fall in from 20 degrees C down to the low teens. One would then think that Elephants prefer to drink at cooler temperatures in order to stay concealed during the hotter times of the day.
With my last samples collected a majority were breeding herds of elephants ranging from 5 members to 25 members. My one worry is that the larger breeding herds may skew the data as the larger numbers make a bigger impact. This could mean that even one fairly large breeding herd coming to drink at a specific temperature may bump the numbers up for that temperature. This could then lead to a misrepresentation of data in the long run, this is something I am keeping in mind whilst writing my discussion.
After some messing around in Excel for a bit I created two figures, which contain two and three graphs respectively, though there is significant overlap between the two. They both contain a graph that displays the relationship between measured sunlight and branch growth, a graph showing the relationship between the distance to the nearest neighbour[ing tree] and branch growth and the second figure includes a graph showing the relationship between distance to nearest neighbour and measured sunlight.
I was surprised at first to see a stronger correlation between distance to “nearest nieghbour” and “branch growth” than between “measured sunlight” and “branch growth”, however this surprise quickly dissipated when the second graph measuring these variables (created from data taken at a second, lower elevation study area) actually showed a negative relationship, while that of “sunlight” and “branch growth” remained positive.
I was also quite surprised to see that the data did not fall into two distinct sets, or groupings, as I had initially predicted would occur (due to the observed distinction between number of branches on the uphill and downhill sides of each tree in the field). Prior to, and during my data collection, there seemed to be an obvious schism between the two sides of nearly every replicate. While the data still show a positive relationship between light measurements and frequency of branch growth, I suspect the sample size was not large enough to reflect this apparent discrepancy I noted in the field.
I know it sounds simple, but I actually struggled a bit with the question of how to properly label the axes. “Sunlight?” “measured sunlight?” “light”? After some deliberation I settled on “sunlight (W/m2)”, though as with which elements I will include in the figures that I put in the final report, I may change this.
I also was unsure about the whether or not to include the third graph (which shows relationship between “distance to nearest neighbour” and “sunlight”). I am still unsure whether both figures will include this graph, or if it will be included as a separate figure in the final report.
I also plan to take a harder look at the captions below each figure when putting together the final report to determine if they require further elaboration.
