| Common species | Systematic sampling percent error | Random sampling percent error | Haphazard sampling percent error |
| Eastern hemlock | 11.5% | 44.5% | 41% |
| Red maple | 4.3% | 8.9% | 22.6% |
| Rare species | Systematic sampling percent error | Random sampling percent error | Haphazard sampling percent error |
| Striped maple | 100% | 4.6% | 4.6% |
| White pine | 90.5% | 100% | 100% |
I compared 3 different sampling strategies using the area-based method for the Mohn hill forest area: random, systematic, and haphazard. Although all 3 strategies had similar estimated sampling times, the fastest was haphazard sampling (estimated to take 12 hours and 29 minutes). The second fastest was systematic sampling (12 hours and 37 minutes) and the slowest was random sampling (12 hours and 43 minutes).
The two most common species were Eastern hemlock and red maple while the two rarest species were white pine and striped maple. The 3 sampling strategies could overall more accurately estimate the density of common species than rare species, which makes sense given they’re more abundant so there’s a larger sample size (less variability) in the spots being sampled. For common species: the lowest percent error estimated was 4.3% using systematic sampling for red maple and the highest percent error was 44.5% using random sampling for Eastern hemlock. Systematic sampling appeared to be the best strategy here. For rare species: the lowest percent error was 4.6% using random or haphazard sampling for striped maple and the highest percent error was 100% using random or haphazard sampling for white pine. The large differences in percent error (and high percent error) between the sampling strategies rare species could be explained by considering their low abundance in the sampling site as a small sample size increases data variability and makes it harder to get a representative sample. Also, the accuracy of the sampling strategies may vary more for rare species from time to time by chance of “picking the right spots” so the instances of low 4.6% error might not be reliably obtained again if the experiment was repeated. Data from 24 sampling sites was not enough to accurately estimate the density of rare species but it was enough to be relatively accurate for common species. This reflects the need to consider factors like the rarity of your subject in the study environment when choosing a sampling strategy.