Tutorial Sampling Strategies and Calculations

In this tutorial, I used area-based sampling methods for all three scenarios. The most efficient sampling method was the systematic sampling, although not by much – it took 12 hours and 5 minutes, only slower by 39 minutes than the next fastest method which was random sampling, coming in at 12 hours and 44 minutes. Random sampling and haphazard sampling took essentially the same amount of time, as haphazard sampling took 12 hours and 49 minutes. For my random sampling simulation, I didn’t end up sampling a single white pine, so the least abundant species ended up being the Striped Maple and Chestnut Oak. The following is the calculated percent error for those respective species within each sampling method.

On average, the rarer a species was found to be for each sampling technique, the higher the percentage error was. For the species that are in higher abundance, the measured value is much closer to the true value, although still with significant error depending on the sampling method. For these abundant species the systematic sampling method was the most accurate, but what surprised me was how inaccurate it was for the rare species, skewing the numbers in such a way that on average, it ended up being the least accurate sampling method by far. Random sampling appears to be the most accurate, but since no White Pine was found by random sampling, the average percent error may be artificially decreased, as the White Pine had the highest percentage error by species sampled. A sample size of 24 may be enough depending on the scope of the project, as regardless of the technique the sampling will still take an entire day, but for more accurate results a higher sample size is probably necessary. There were not many cases of the percentage error being below 10%, and in several cases the error was so high the measured data may be considered useless.