I choose Snyder-Middleswarth Natural Area in the virtual tutorial. 25 plots are sampled for each technique.
Among three sampling techniques,
Haphazard: 12h 59min
Random: 13h 14min
Systematic: 12h 34 min
Systematic sampling had the fastest estimated sampling time.
2 most common species are Eastern Hemlock and Sweet Birch, percentage errors are calculated below:
Haphazard | Random | Systematic | |
Eastern Hemlock | (468-469.9)/469.9*100%= 0.4% | (440-469.9)/469.9*100%= 6.4% | (360-469.9)/469.9*100%= 23.4% |
Sweet Birch | (88-117.5)/117.5*100%= 25.1% | (148-117.5)/117.5*100%= 26% | (124-117.5)/117.5*100%= 5.5% |
2 rarest species are White Pine and Striped Maple, percentage errors are calculated below:
Haphazard | Random | Systematic | |
White Pine | (4-8.4)/8.4*100%= 52.4% | (0-8.4)/8.4*100%= 100% | (12-8.4)/8.4*100%= 42.9% |
Striped Maple | (0-17.5)/17.5*100%= 100% | (16-17.5)/17.5*100%= 8.6% | (20-17.5)/17.5*100%= 14.3% |
By comparing the percentage error of the different strategies for the two most common and two rarest species, the accuracy changes with species abundance. The accuracy is more likely to decrease when the species are rarer.
Overall, when calculating for the most common species, Haphazard is more accurate than other two strategies, however, when comes to the rarest species, the systematic strategy is more accurate than haphazard and random strategies.
In sum, to minimize the error, I think obtaining larger sample size are important, and different strategies may be suitable under different circumstances, so it should be choose wisely when establishing the experiments.
Looks good and hopefully it helps you think about your own sampling.