I used the distance based method for the virtual forests tutorial, and sampled the trees using the three sample placement techniques.
Below is a table showing the percentage error for different sampling techniques.
Table: Percent Error
| Systematic Sampling | Random Sampling | Haphazard Sampling | |
| Eastern Hemlock (common) | 13.2 % | 7.0% | 93.1% |
| Sweet Birch (common) | 16.7% | 21.3% | 62.6% |
| Yellow Birch (common) | 17.5% | 46.0% | 46.2% |
| Chestnut Oak(common) | 4.5% | 24.8% | 9.0 % |
| Red Maple(common) | 31.3% | 1.1% | 60.7% |
| Striped Maple (Rare) | 40.0% | 4.0% | 40.0% |
| White PIne (Rare) | 100% | 1.1% | 89.2% |
| Estimated time to sample | 4 hours 16 minutes | 5 hours 1 minutes | 4 hours 40 minutes |
Systematic sampling was the most time efficient with an estimated time of 4 hours and 16 minutes.
There was a big fluctuation in the accuracy of the sampling strategies in the common species. Random sampling was the most accurate for the most common species (Easter Hemlock). Red Maple also had the most accuracy using random sampling, however the rest of the common species displayed the more accuracy using systematic sampling.
For rare species, random sampling showed the most accuracy.
The accuracy declined greatly for rare species.
The difference in the accuracy, and the very low density of rare species suggests that 24 sample points is not enough to accurately estimate the abundance of these species. Having more sample points will increase the accuracy as it will give us more data to work with.