Category: Post 4: Sampling Strategies

For this assignment, I used the area systematic, random, and haphazard sampling strategies in the virtual forest tutorial. The technique with the fasted estimated sampling time was the area systematic sampling strategy (12 hours, 5 minutes). I compared the percentage error of the different strategies to the two most common species (Eastern hemlock & sweet birch) and two rarest species (Striped maple & white pine).

I found that as the species abundance increased so did the percentage error. For example, in the systematic area sampling method 228 species were counted and there was a 1.6 percentage of error for Eastern hemlock. Comparing this to the haphazard area sampling method, which had a species abundance of 297 samples, the Eastern hemlock percentage of error was 39.2%.

It appears the random sampling strategy was the most accurate. Below is a summary of the data collected during the three sampling strategies I tested.

Area – Systematic: Area sampled 2400m2, 24 samples

Two most common species:
Eastern hemlock – Density 462.5 –1.6% error
Sweet birch – Density 170.8 – 45.4% error

Two rarest species
Striped maple – Density 8.3 – 52.6% error
White pine – Density 4.2 – 50% error

Species abundance: total species sampled 228

This sampling strategy had the Fastest estimated sampling time

Area – Random: Area sampled 2400m2, 24 samples

Two most common species:
Eastern hemlock – Density 462.5 –1.6% error
Sweet birch – 158.3 – 34.7% error

Two rarest species
Striped maple – Density 4.2 – 76% error
White pine – Density 12.5 – 48.8% error

Species abundance: total species sampled 216

Area – Haphazard: Area sampled 2400m2, 24 samples

Two most common species:
Eastern hemlock – Density 654.2 –39.2% error
Sweet birch – Density 120.8 – 2.8% error

Two rarest species
Striped maple – Density 50.0 – 185.7% error
White pine – Density 8.3 – 1.2% error

Species abundance: total species sampled 297

Blog Post 4

Which technique is the most efficient in terms of time spent sampling?

Systematic: Sampling along a topographic gradient 4 hours, 5 minutes (20 minutes faster than haphazard)

This is from a distance based sampling strategy.

Percent error

Systematic: Sampling along a topographic gradient

Percent error common species (Eastern Hemlock): 12.9%

Percent error rare species (Striped Maple): 46.9%

Random: Distance, random or systematic

Percent error common species (Eastern Hemlock): 13.6%

Percent error rare species (Striped Maple): 58.9%

Haphazard or subjective sampling:

Percent error common species (Eastern Hemlock): 6.5%

Percent error rare species (Striped Maple): 100%

 

The most accurate sample strategy for common species was Haphazard or subjective sampling.

The most accurate sample strategy for rare species was Systematic: Sampling along a topographic gradient.

The accuracy for rare species declined over the sampling methods used.

The accuracy rate changed in relation to species abundance.  The less abundant a species was correlated with a greater increase in error of collecting a sample for that species.

24 sample points was not enough points to capture the number of species in this community. In the Haphazard sample, the error rate of 100% would miss this species entirely. The sample strategy that most captured this species still had an error rate of 46.9%, which greatly under-represents this species.  Adding more sample locations would reduce the error rate.

The technique with the fastest sampling time was the random sampling method. The Red Maple and the Eastern Hemlock were the two most common species and the Striped Maple and White Pine were the two rarest. The percentage error for these with the different sampling techniques are as follows;

 

	Systematic	Random	Subjective
Eastern Hemlock	11.5%	21.7%	10.8%
Red Maple	4.1%	13.4%	30%
Striped Maple	??	347%	42.9%
White Pine	47.5%	??	??

?? denotes the PE was unable to be calculated because the species was not found in sampling.

 

The accuracy of the sampling appeared to increase with increased species abundance. The rare species had much higher PE values than the common species and some were unable to be calculated at all because the species was not found using certain sampling techniques.  The Random and Systematic sampling techniques were on average about 4-5% more accurate than the Subjective/Haphazard method of sampling for common species. However, when calculating rare species the Systematic and Haphazard method were much more accurate than the Random method, but due to the lack of species abundance and data it is difficult to read too much into this.

In the virtual forests tutorial, the three sampling strategies used were Systemic, Random and Haphazard with an area-based approach. According to the results, the random sampling technique had the fasted estimated sampling time of 4 hours and 45 minutes. The other two sampling methods, systemic and haphazard sampling had a much slower estimated sampling time of 12 hours and 35 minutes and 12 hours and 44 minutes respectively. In this tutorial the most common species were Eastern Hemlock and Red Maple. The calculated percent error for Eastern Hemlock and Red Maple for each of the sampling methods were:

• Systematic
  • Eastern Hemlock – 6.36%
  • Red Maple – 24.47%
• Random
  • Eastern Hemlock – 23.26%
  • Red Maple – 42.72%
• Haphazard
  • Eastern Hemlock – 8.07%
  • Red Maple – 12.53%

In contrast, the two rarest species were White Pine and Stripped Maple. Their calculated percent error for each of the sampling site were:

• Systematic
  • White Pine – 233.33%
  • Stripped Maple – 100.00%
• Random
  • White Pine – 100.00%
  • Stripped Maple – 21.14%
• Haphazard
  • White Pine – 138.10%
  • Stripped Maple – 31.43%

Based on these results, it can be concluded that the Haphazard or Systematic sampling methods were the most accurate for the most common species and the Random or Haphazard techniques were the most accurate for the two rarest species. However, as demonstrated by the above results, the accuracy of the sampling methods declined when looking at the rarest species. The high variation between the percent error values could be a result of the small sample points. 24 was not a sufficient number of sample points; in order to produce more a accurate estimation of abundance  more sample points is necessary.

In the virtual forest tutorial I used the area based sampling method for all three sampling strategies. The technique from the fastest estimated sampling time to the slowest estimated sampling time was systematic sampling technique at 12 hours and 7 minutes, haphazard at 12 hours and 27 minutes and random sampling at 12 hours and 47 minutes. The 2 most common species in the Snyder-Middleswarth Natural Area were Eastern Hemlock and Red Maple. The 2 rarest species were Striped Maple and White Pine. The percentage error for Eastern Hemlock for systematic, random and haphazard sampling were respectively: 3.34%, 4.23%, and 8.17%. The percentage error for Red Maple for systematic, random and haphazard sampling were respectively: 5.38%, 40.45%, and 19.42%. The percentage error for Striped Maple for systematic, random and haphazard sampling were respectively: 90.29%, 18.86%, and 28.57%. The percentage error for White Pine for systematic, random and haphazard sampling were respectively: 100%, 50% and 296.42%. The accuracy changed with species abundance. For the two most common species the sampling strategy that was more accurate was systematic. For the two rarest species the sampling strategy that was more accurate was random sampling. The accuracy declined for rare species. The large percentage error may be because 24 was not a sufficient number of sample points to capture the number of species in this community or to accurately estimate the abundance of these species.