|
| |
Wlf 543
E. O. Garton
Sequential Sampling
Problem Set No. 4
Many national forests in Oregon, such as the Deschutes
National Forest, have set standards for snag management which require
maintaining 0.9 snags per acre (2.2 snags per hectare) well distributed
throughout the watershed. Bate,
Garton and Wisdom (1999) developed a variable width strip-census approach to
sampling snags that can be applied to stands to assess snag abundance.
They recommend sampling 50 m long plots which are 10, 20, 30 or 40m in
width depending on the density of snags and other vegetation in the area.
- Develop
a sequential sampling scheme which could be used to sample stands following
harvests to insure that the stands met this standard. Your goal is to be 90%
certain that the stand has at least 2.2 snags per hectare.
Assume that you want to detect stands with less than this abundance
of snags at least 80% of the time. Any stand with more than 3 snags per
hectare is considered to be in excellent condition.
[Note: Assume a normal
distribution of counts but a Poisson-like pattern in which the variance is
equal to the mean.]
- Design
a sequential sampling scheme to insure that the snags are well distributed
throughout the stand also. Well
distributed will be defined as requiring that on average the proportion of
the 40m x 50m plots that has a snag present is no less than 44%.
If this percentage is greater than 60% the snags are extremely well
distributed.
- Develop
a QuattroPro or Excel spreadsheet to perform these calculations and and plot
the sequential sampling graph for making decisions.
|