Category: Post 6: Data Collection

Blog Post 6 – 16/02/20

Field collection went well today. I went out to collect data at 1700 hours. The skies were clear and the temperature was mild, a few degrees above zero. There was a slight wind and some mild snow cover, but otherwise there were no issues with collecting my data. Today I surveyed the location of the trees within the randomly selected sample locations. The tree sampling went well and there were no issues with implementing the stratified random sampling design. Firstly, I divided the areas into the three strata of pond land, central park land, and edge land. 15 transects were then randomly selected with 5 quadrats per transect. Transect amounts reflected the percentage covered by each strata namely 8 transects for central park land, 4 transects for edge land, and 3 transects for pond land. I then marked each quadrat with an “X” to indicate the presence or absence of each of the three tree species, white spruce (Picea glauca), aspen poplar (Populus tremuloides), and white birch (Betula papyrufera). Upon collecting this data I noticed that there was a surprisingly high amount of white birch trees in the pond land transects, which was unexpected and caused me to reflect upon my hypothesis. I am now wondering if the white birch (B. papyrufera) tree species requires a higher soil moisture level to survive. However, a substantial amount of white birch was also found in the central park and edge land transects, therefore it is possible that the which birch thrives in all soil types. I will do some research on the growing conditions for white birch and use the information to reflect upon my findings. For this data collection I also sampled three replicates from each of the three strata to determine the soil moisture content. As expected, there were higher levels of moisture in pond land soil than anywhere else in the park, while edge land soil had the lowest levels of moisture. In essence, the moisture sampling also went very well despite the light snow cover and overall the data collection went very well today. 

Over the past weeks I’ve visited the site a few more times and have indeed noticed additional ancillary patterns in regards to the presence of Alnus rubra. The remaining trees have a more clay-based soil still present around them and this suggests their survival is potentially related to the edaphic conditions of the soil that have not washed away yet. I have also recognized mistakes in my initial experimental design, and attempted data collection from a distance (as noted in one of my last posts) but it is not possible. It’s because of these factors that I need to adjust my hypothesis to reflect the new patterns i’ve noticed and I also need to alter my variables and sampling technique.

As the shoreline continues to rapidly erode at this location, the loss of soils capable of providing growing area for Alnus continues to disappear. The soil that continues to be washed away (mostly clay-based) appears to be the desired location for whatever Alnus rubra are present and potentially the desired location for their seed. These edaphic changes, combined with interspecific competition from plant species that are better adapted to the more sandy or gravelly soil mean that Alnus rubra will eventually no longer be able to survive and reproduce in this area.

If I believe that Alnus’ presence will decrease as soil type moves from clay to gravel, but may also depend on the plants that are most dominant in these zones as well, I can record data for these using two separate models and then analyze them to see if plant dominance does affect Alnus presence.

Response variable 1: Presence of Alnus rubra

Predictor variable: Soil type (stony/gravel zone, sandy/gravel zone, clay mix zone)

Response variable 2: Dominant plant type (based on plant with most % cover in each zone)

Predictor variable: Soil type (stony/gravel zone, sandy/gravel zone, clay mix zone)

Sample unit: 400-500 meter transect across the shore. The width is 6 meters, going up from the high tide line to the level top of the slope.

Subsample unit: Quadrats are randomly generated # of paced meters, continuing from one quadrat to the next. This area of study is difficult to access and is hard to record accurate data unless I use bigger quadrats than my last attempt. Each quadrat is a 2m (along shore) by 6m (up slope) rectangle. The presence of Alnus in each soil type as well as the dominant plant species in each soil type is recorded.

40 quadrats were randomly placed along the transect and data was recorded. A second round of data collection with a new set of random quadrats was completed on the same day, therefore I had one replicate.

I am expecting continued adjustments to my hypothesis and design as data is recorded and analyzed.

I have collected 30 replicates divided evenly among three transects approximately 100 meters in length. The transects were selected using systemic sampling consisting of one for each direction; north, south, east, and west. The replicates were selected using random sampling by using a stop watch to randomly produce a number to determine how many steps to take to find my next replicate. Each replicate was evaluated for trunk diameter at chest height and aspect of growth. Collecting sample data has been difficult as of late due to inclement weather including deep snow and extreme cold. A pattern noted in my data is that Douglas fir trees found on an eastern aspect had the smallest average trunk circumference, followed by trees on a northern aspect, trees on a southern aspect, and trees on a western aspect with the greatest average trunk circumference. In the context of my hypothesis, this data has caused me to reflect on my prediction that trees growing on a northern aspect will have a larger average trunk circumference.

Create a blog post describing your field data collection activities. How many replicates did you sample? Have you had any problems implementing your sampling design? Have you noticed any ancillary patterns that make you reflect on your hypothesis?

I have implemented three data collection activities to recognize the healthiness of ecology depending on humans intervention rate.

First observation  was collecting the ratio of fresh vegetation per area. Fresh vegetation proportion was measured by the colour, green. In a quadrat (10cm*10cm), if the area is composed of greens more than 50%, it would be recorded as 1. If the area observed had green vegetation less than 50% of the area,  it was measured as 0. Each landscape, ornamental garden, ornamental steps, and preserved hill side, was measured 8 quadrats and it was measured with 5 replication. Since the area wasn’t that big, the quadrat selection might overlap easily. I had to use systematic sampling techniques. The area was divided into 5 area and quadrates were selected in subdivided area per visit. After one visit the other area was selected for data collection. In this way, overlapping of quadrat was avoided.

Second observation was collecting the number of vegetation species observed per quadrat (10cm*10cm). As above the area was divided in to five and subdivided area was observed each visit. 8 quadrats were observed per visit and was measured with 5 replication.

As going through both experiment 1 and 2, there should have been consideration of the amount of water the land received and type of soil the vegetation grows. They both affect highly in vegetation growth. The hill nearby the church isn’t actively managed by someone, the amount of water the landscape receives and type of soil they grow on might be different. On the next observation, observing this point is another important criteria.

The third observation was measuring the bird activity rate depending on landscape. Bird activity should be considered morning, afternoon and evening. The bird activity rate changes during the time of the day. Therefore, each visit per day must be three times; morning (9am-9:30am), afternoon and evening, with 10 replication. Each observation lasted 10 minutes and the number of birds flying around the landscape was measured.

Background: I have decided to change my hypothesis and organism I am studying to make my sampling simpler. I have observed that deer fern are more present in some areas of Burnaby Mountain covered in forested than areas than areas that don’t have trees. To keep this simple I will categorize a sampling area as either shaded having most of the sky covered by the tree canopy, partially shaded having part of the sky shaded and non-shaded. These three zones represent differing levels of sunlight that is able to reach the ferns on the ground. My new hypothesis is that the deer ferns are more successful under lower lighting levels. As such my prediction is that I will find a greater abundance of ferns in shaded and partially shaded areas than non-shaded areas.

I chose three sites to conduct my sampling each had a non-shaded area, partially shaded area and shaded area. At each area I recorded 10 samples. For a total of 90 replicates. My sampling unit was 1m2. Which I had planned to measure the ground cover of the ferns; however, I found that most ferns covered the whole area. So instead of trying to measure ground cover I just recorded either the area being mostly covered by deer fern, grass, bare ground, tree or thorns. Of my three sites one of the non-shaded areas is managed to some extent because a gas line is below so they clear cut the area every few months. I think this has had the effect of reducing the amount of ferns that can grow because they are in competition with other plants that perform better under higher lighting conditions. The partially shaded zones had the most ferns with 15, shaded at 5 and no-shade at 6. This is what I had initially observed; however, I had expected to find more ferns in the shaded areas. I observed in the shaded areas underneath the trees (evergreens) that very little grew, the ground was mostly covered in pine needles. In the partially shaded areas the trees were mostly thin in diameter deciduous trees. One of the non-shaded areas had a high number of ferns which was the opposite of the other 2 non shaded areas that had only 1 fern each.

During my field data collection, I collected 10 replicates from each of my two substrates, access to sunlight (uncovered from tree cover) and limited access to sunlight (covered by tree cover), this amounted to 20 replicates in total. I collected my replicates using a 1.0 by 0.5 meter quadrat and collected them along a transect in a straight line and randomized my steps between each sample. The quadrat was placed beside where my foot fell on the last step. After my initial collection, I had considered doing a point count rather than a percent cover but decided to stick with my initial measurement method of percent cover as the vast majority of the cover was English Ivy with few other plants to have to work around so I decided I will have adequate accuracy using percent cover.  I did not face many challenged when collecting data, besides having to be very careful walking around the Ivy as it got very thick in some areas and I did not want to harm them! I did not notice any patterns that made me reflect on the hypothesis. My two substrate areas are very close in proximity, the major difference I observed between them being the tree line where the forest canopy starts and stops but it creates a drastic difference in the abundance of the Ivy.