Post 4: Sampling Strategies

Three sampling methods were used to capture the relative density of different tree species in a forest community.

1. Sampling along a topographic gradient proved for the most consistent results a crossed the species from highest density to the low density. The hesitation with this style of sampling though is that if the gradient is meant to be representing a relatively large plot of land, one linear stretch of data wont due justice to the remainder of the space. Several linear gradients a crossed the community being sampled would lessen the percentage error.

Density Percentage Error Calculations

Eastern Hemlock 30.2%

Sweet Birch 18.3%

Striped Maple 37.1%

White Pine 4.76%

2. Random Sampling proved for more accurate results within the most common species, however in my randomized sampling attempt I found 100% error within the two rarest species. Therefore, this style of sampling has the potential to omit species due to the fact that it cannot take into account any topographic changes or tendencies for abundance of certain species in certain areas, which may be require when trying to quantify a community.

Density Percentage Error Calculations

Eastern Hemlock 28%

Sweet Birch 11.5%

Striped Maple 100%

White Pine 100%


3. Haphazard sampling in my attempt was done with the intent on evenly spacing the 5 plots within each subsection along the y axis, in a sort of grid like plot in an attempt to gain the most accurate representation of the forest community. My percentage error results do indicate the most accurate results of the three sampling methods however with a percentage error at above 50% I still feel like more sample plots should be taken then 24 if this is the result.

Density Percentage Error Calculations

Eastern Hemlock 8.3%

Sweet Birch 11.1%

Striped Maple 54.0%

White Pine 52.1%



Sampling along a gradient is the quickest sampling method, estimated at 12 hours and 37 minutes, while the other two estimations were identical at 13 hours and 10 minutes. Sampling along a gradient was also the only method which didn’t return a percentage error calculation of over 50 % which deems it to be relatively accurate in terms of quantification of abundance of species in my experiment. This however is subject to change if the topographic gradient that was selected is a poor representation of the remainder of the community. With the calculated percentage errors of the abundant and least abundant species, it is my opinion that more than 24 sample plots would be required to minimize the ambiguity of the forest community.

Leave a Reply

Your email address will not be published. Required fields are marked *