Using the virtual forest, I chose to sample the Mohn Mill site, using random, systematic, and haphazard sampling methods, each with 25 samples (oops). The sampling method with the fastest sampling time was systematic sampling (12h, 36m), working along an east-west transect and sampling at set quadrats. Random sampling was the slowest methodology (13h, 40m), followed by haphazard sampling (13h, 2m).
I compared percent error (PE) for the 3 most common, 3 occasional, and the 3 rarest species, using R for analysis and visualization. This showed that random sampling had the lowest overall PE, while systematic sampling had the highest for rare and occasional species. Rare species had low percent error (with one outlier – Yellow Birch = 900%) but that is due to a low detection of these species with any of the sampling methods. Haphazard sampling had the highest PE for occasional species. Random sampling was the most accurate method overall. Haphazard sampling significantly overestimated occasional species. Accuracy of all methods increased with abundance.
Depending on study objectives, additional sampling points may be necessary. For example, if the study intended to characterize the dominant plant communities, and used the random sampling method, 25 sample points would be suitable. This was not enough sample points to accurately estimate abundance of occasional or rare species.
yes good interpretation