The virtual forest sampling exercise produced some interesting results.
Sampling Time – In order from least to greatest: Systematic (12:36), Random (12:41), Haphazard/subjective (12:54). While this makes sense from a logistical standpoint in terms of total distance travelled, travel distance between points and time spent deciding where to sample, I suspect that that the differences in time would be much greater if I tried to carry out this exercise in real life. If systematic sampling via walking in a straight line through the forest is used as the baseline, I’d think that trekking all over the forest between random points while trying to locate the gps coordinates without any consideration of the terrain would add significantly more than 5 minutes to the sampling period. I also suspect that the subjective sampling would be somewhere in the middle, because I would naturally take the easiest path through the forest and only sample quadrats that were easy to get to, which would not necessarily be the case for the random sampling method.
Percent Error – The percent error generally increased with the rarity of the species. The systematic method was generally the most accurate overall, and random the least. The subjective method was the most accurate of the three for rare species, this could be because I chose to place quadrats fairly evenly spaced within each strata, however chose “representative” looking communities. I suspect this would be the case in a real subjective sample as well, as humans would naturally be drawn to new and interesting specimens when deciding where to sample.
