User: | Open Learning Faculty Member:
I took my initial samples along Dallas Road and Beacon Hill Park (Victoria, BC) this week. The weather was sunny, 8°C and windy. I used a 0.3m x 0.3m quadrat, systematic sampling strategy and counted the number of Nootka dead and alive rose hips at three locations along a gradient differing in distance from the ocean. When I arrived at the first site (an exposed bush along a coastal bluff near Dallas Road), I was surprised that nearly all the rose hips were dead. I observed that the coastal bluff had a higher proportion of red, alive hips earlier this month. It appears the winter has been hard on the Nootka Rose.
I was nervous that the other sites would also display a similar trend; however, the sites further from the ocean, an intermediate site called “deciduous forest” and the end range site called “Beacon Hill” did have some alive hips.
I came across a difficulty using my quadrat as I placed the quadrat on the top of the bush, and there would be hips present at different depths beneath the quadrat. Therefore, I will have to define my sampling area as within the same plane as the quadrat.
As well, I will need to find more sites within each gradient for the final data. There are not enough sample units from the single site I collected data from for the initial data. I have confidence that there are other sites within each gradient where the Nootka Rose grows.
I will continue to collect data in the same manner; however, modify my definition of the quadrat region and search for additional sites within each gradient. Redefining the area of the quadrat region will make the replicates more standardized. As well, collecting more data from additional sites will provide a more accurate measure.
Interesting subject, your sampling method is described really well. I came across a similar problem, as I also found that five sites was not giving me sufficient data to generate accurate results. I wonder if that was the case for most students. Best of luck in your search for other quadrat regions,
Hissan
take a look at the ‘rule of 10’ in the modules.
for fun, google ‘how many samples do I need’, and you will see relationships between sample numbers and certainty. Very often we trade off resources (time and sampling ability) with what we feel is a reasonable amount of sampling