Effective sample size

[ecolo1]

Hello,

I have 1000's of data points estimating coverage of the amount of detritus on a forest floor using quadrats.

I want to explore the number of samples needed to get a stable estimate of the mean from this population. e.g. what sampling effort from this entire population would be required to see the variance stabilize to an acceptable level?

Any ideas, including programs to use.
Many thanks :)

[quark]

You may want to determine which sampling scheme to use. Simple random sampling is the most commonly used, and you can also use stratified sampling, cluster sampling, systematic sampling, etc. Each scheme will require a specific sample size.