Blog Post #4 – Sampling Strategies

For the Virtual Forests tutorial, I chose to use the area-based methods for my 3 samples. The fastest technique for sampling was the systematic technique along a topographic gradient with a time for 12 hours and 36 minutes.  What surprised me about the results was that the random and haphazard techniques, each taking 13 hours and 14 minutes, did not take much more time than the systematic approach.

The two most common species I found in my samples were the Eastern Hemlock and Sweet Birch.

Analysis of the data collected for the 2 most common species indicates that the haphazard method of sampling was the most accurate strategy, with both common species having percentage errors in the single digits.

Analysis of data collected for the 2 most rare species shows that all 3 sampling methods provided very inaccurate results. The second rarest species, the Striped Maple, was well sampled in the random method with a percentage error of only 8.6%. However, the rarest species, the White Pine was not found at all using this method. As species abundance decreased, percentage error of sampling using all 3 methods decreased.

Overall, 24 plots does not appear to be enough to get an accurate representation of species density across the range of species in the geographical area. I would predict that increasing the number of plots would increase the accuracy of all 3 sampling techniques.

To test this theory, I repeated both the haphazard and systematic techniques using 50 plots instead of 24. I found the same number of species (7) as before, however the haphazard method now yielded percentage errors of 0.02% for the most common species (Eastern Hemlock) and 19% for the rarest species (White pine). The systematic method now yielded a percentage error of 3% for the most common, and 19% for the rarest species.  I conclude, based on this observation, that more sampling plots, regardless of method, yield more accurate results.