Category: Post 5: Design Reflections

Blog Post 5 – Design Reflections

TRU Open Learning Leave a comment

Difficulties in executing the sampling strategy was mostly due to recognition. Since I’m not an ornithologist, I cannot say for certain that I did not count the same bird more than once.

The data that was collected was not overly surprising. I will continue collecting the data for the different birds using the same method however two locations were added, the science building and house 9. As well a couple of observations were made in each location to count the number of people to back up the predictions of which area has the most human traffic within a day.

Blog Post 5: Design Reflections

TRU Open Learning Leave a comment

The only difficulty I encountered was when I originally when to my study site to collect data and the edges of the river hadn’t been completely unthawed. Considering I would be measuring the water depth that cattail was growing, I couldn’t exactly measure frozen water. To fix this problem I just waited another two weeks until the water was completely thawed. The data I collected was not at all surprising and I don’t think my approach needs any modification.

Post Five: Design Reflections: Cates Park

TRU Open Learning Leave a comment

My sampling strategy had a few difficulties, and therefore I decided to attempt another, hoping to redeem my first effort.

The first sampling strategy used a transect with alternating quadrats. Using my roommate’s measuring tape was the first challenge, since it only had imperial measurements, so I had to convert data into centimetres. I’m grateful I had a willing assistant who could help lay the measuring tape along the necessary gradients. The data collected was surprising as it revealed low numbers, and I realized that my next similar attempt should be on a more grand scale. I will need to be creative with data collection along points that are steep or heavily forested. One other difficulty was creating a data sheet template that would work for my purposes. I improvised and moved the data to a new spreadsheet that was more organized.

The second set of data collected was haphazard and distance based, and I believe, more successful. Five trees were selected haphazardly for ease of access in this forested region. These were the centre point where I measured neighbouring species. Again, the tape measure was not an ideal tool, and I benefited from having someone to assist. After the data collection, I realized I should have created a map, image or layout of where each tree was situated in relation to the midpoint. This data was predicted but I’m looking forward to more sampling.

I will likely continue to collect data with the second approach, and add another kind of sampling strategy to assist in the bigger picture of my hypothesis. By adding varied sampling techniques, replicates and variables, I will likely be able to prove or disprove my prediction and hypothesis. Modifications to data collection will also include appropriate measuring techniques and recruiting more volunteers!

 

Blog Post 5: Design Reflections

TRU Open Learning Leave a comment

While collecting the data, some difficulties faced were the number of people on the trail and attempting to identify the number of individuals for each species. On the first point, when making observations there were times when tour groups of children would approach the area to observe birds. Although they attempted to remain silent, the sheer number of people in one location made it that the birds were not nearing the area. Approximately five minutes after the groups would leave, the birds would return to this area. To address this issue, my sampling design will be modified such that if there are touring groups in the area, I would wait five minutes after the group has left before beginning the data collection.

On the second point of identifying the number of individuals for each species, unless the individuals are tagged it is challenging to determine the exact number of individuals there are for each species. To address this issue, I decided that species diversity would be assessed for this study rather than abundance.

The data collected was slightly surprising in that the abundance of individuals for each species was greater in colder weather. The diversity of species was similar in colder temperature as compared to warmer temperatures. Despite these observations, only five replicates have been collected thus far, it will interesting to see how/if things change.

Post 5: Design Reflections

TRU Open Learning Leave a comment

Kevin Ostapowich
April 1, 2019

My study is looking at the distribution of birch within a mixed forest.  Using QGIS, I overlayed a 30m x 30m grid over my study area and haphazardly chose 5 sites to count individual trees.  I then went into the field, using GPS to find the sites, and counted trees of all species within the plots.  I didn’t have too much trouble implementing this procedure but I noticed that the different tree species tend to be grouped in clusters which may have skewed the results.  For example, one of my sample plots was almost entirely composed of pine while the surrounding forest was deciduous.  Another  plot was almost entirely aspen with no birch but immediately surrounding the sample area was a lot of birch.  The plots that I counted may not be entirely representative of the forest composition.

There are two solutions that I can see:  choose a larger grid size to encompass a greater diversity of tree species, or sample more locations.  For consistency and to keep sampling manageable, I plan on continuing with the method that I have already used and to increase the number of sampling sites.

Post 5: Design Reflection

TRU Open Learning Leave a comment

There were some difficulties in implementing your sampling strategy. The difficulties mainly surrounded what parameters to record when examining the absence or presence of English Ivy. Such difficulties were if the type of tree or what side the ivy was growing on should be recorded. I wanted to keep the sampling strategy straightforward without generalizing.

I plan on continuing using the same data collection technique as it is straightforward and efficient to using randomly selected point-radii to examine the absence or presence of ivy.

Blog post 5

TRU Open Learning Leave a comment

The collection of data for my project proved to have some challenges. Due to the timeframe of my sampling, the dirt samples were hard to collect with the years first real snow dump falling down. The two plots I was sampling also proved to be quite rocky, so it took some time to find spots I could dig to an acceptable depth as to have a soil core that would be indicative of the actual soil pH. When I was preparing the samples for testing with a pH meter I had to pick out as much of the organic matter and gravel as possible which also proved challenging. Once the samples were separated down to just dirt, I used the pH meter get some actual numbers, but the pH read was rather uniform in its distribution across the samples, hovering at around 6 pH. Even the control sample of just distilled water read a surprisingly low pH level of 5.75.  I definitely need to re-approach how I’m going to gather my data; however, I’m relatively sure that the results are due to my methods used rather than the actual samples. I found online that most studies use a salt dilution to help the sensor pick up the pH, as in distilled water there may be too few ions for the sensor to properly pick up the change in pH from one sample to the next. The other solution is to use a more even distribution as some researchers just use a 1:1 ratio for water to soil mix it up and give it 5 minutes to settle. I also plan on acquiring a sieve to help remove the remaining organic matter that I couldn’t grab and double check the results of the pH meter with pH strips. I think with the sieve and using a closer ratio of dirt to water the actual pH should be easier to pick up, and the pH strips will allow me to double check my results.

Blog Post 5 – Design Reflections

TRU Open Learning Leave a comment

During my initial field data collection efforts at Terwillegar Park, I had no major difficulties implementing the stratified sampling strategy. However, the one difficulty I started to encounter was an increase presence of people and dogs in the park by the time I was finishing up the last point count locations. Due the increased presence of people and dogs present I was encountering disturbance during my 10 minutes of counting birds. Due too this disturbance I have opted to decrease the amount of time spent at each point count location to 5 minutes. I don’t anticipate that decreasing the time spent at each site will have a large effect on the study as I was noticing that I was counting most birds present within the first 5 minutes. Decreasing the time spent will allow me to ensure that each point count location is surveyed with no additional disturbance at the sites later in the morning.

Blog Post 5

TRU Open Learning 3 Comments

For my study, I am examining whether the density of ash trees (total number of trees) in a 25X25  foot plot is correlated the rate of infestation of emerald ash borer. To measure this I initially looked whether there are serpentine galleries in the bark of the trees or if there are fissures in the bark. I noted the total number of trees in each plot and the total number of trees that showed evidence of infestation.

Based on my initial data one issue I had was ensuring not to double count trees within the 5 plots which I was observing ash trees. I overcame this challenge by marking trees with pink chalk. Another challenge I had when looking at the bark for serpentine galleries or fissures was that the ash trees are quite tall so it is hard to notice either of these signs of emerald ash borer presence up top in the canopy (even with binoculars). Additionally, on some trees I noticed serpentine trails in a small area of the tree where the bark had fallen off, this made me wonder just how many of the other trees had serpentine trails underneath the bark that I was unable to observe. Based on this I suspect my initial results on how infected ash trees in the plots were by emerald ash borer could be significantly lower than the reality. I will revisit the same plots as last time (I noted the GPS coordinates of my five 25 X 25 foot plots), however, I will also be taking note if there are other factors of emerald ash borer presence to get a better diagnostic of the health of the ash trees. For example, I will look to see if there are D-shaped emerald ash borer exit holes in the bark or if there are signs of woodpecker activity as well.

Post 5: Design Reflections

TRU Open Learning Leave a comment

I used a random sample technique to choose the 4 sites I selected. In order to keep it truly random, I threw a stick randomly and wherever that stick landed would be the area for the new plot. Once the plot was selected, I used a measuring tape to measure a 10 m x 10 m plot. A wooden steak was placed at each corner and then string was wrapped around all 4 wooden steaks of the plot so that I could clearly see the outline of the plot. Once the quadrat was set up, I walked in transects throughout the plot and recorded the data. A main difficulty I had implementing this was that it was hard to tell which trees, especially dead fallen trees, were spruce trees and not another species. I did some research to find out that in order to tell if the tree has died due to the spruce beetle, that I should look for patterned lines and wholes on the bark as that’s a sign that that tree has died from the beetles. Another difficulty is that in the denser tree cover plot, it was hard to count all the trees without losing count since there were so many trees. To try to fix this, I just made sure that I recounted the plot 2 or 3 times to ensure I have the correct number. Something I found surprising is that plot 2 (with sparse tree cover), actually had a higher amount of dead trees than the other plots. I think the technique I used works well for the area that it is and that by throwing the sticks, ensures that the plots are random.