For the Virtual forest sampling tutorial I chose 3 different sampling strategies: Systematic (along a topographical gradient), Random Sampling, and Haphazard (subjectively without bias).
The technique that had the fastest overall estimated sampling time was:
Haphazard (12 hrs 33mins) followed by Systematic (12hrs 35mins) and then Random (12hrs 46mins). All very close in estimated sampling time, with the largest difference being 13 minutes.
Percentage Error for the two most common species:
Most Common Species: Eastern Hemlock followed by Sweet Birch
- Systematic: 2.1% , 5.5%
- Random: 24.6%, 17%
- Haphazard: 3.7%, 2.8%
Two Rarest species: White Pine followed by Striped Maple
- Systematic: 100%, 8.6%
- Random: 50%, 114%
- Haphazard: 1.2%, 76%
Accuracy most definitely changed with species abundance. With the more common species percent error was much lower (especially with Systematic and Haphazard sampling strategies) than with the more rare species.
I believe in this instance Systematic Topographical gradient sampling strategy was the most accurate, followed by Haphazard Sampling then by Random Sampling.