why 20 many plots pr stratum

[Frederik.h.kallesoe]

Hi everyone

I am currently planning to use itree eco to do estimate the pollution mitigation for 50 small woodlands within a city 0,5-1 ha of size. Here my plan was to run an eco model for each forest with a sampling design, however this would never bring me close to the 200 plots suggested in itree, or the 20 plots suggested for each stratum. However I also expect my woodlands to be fairly homogeneous, and much more that random plots within a city. Therefore I wanted to ask to what extent the minimum sample size is due to the model mechanics or an assumption about the heterogeneity of the city landscape?

Ps I did consider full inventories of each forest, however with limited manpower this is simply not realistic.

Kind Regards,
Frederik Kallesøe.

[Jason.Henning]

Hi Frederik,
The recommended sampling sizes are entirely based on the heterogeneity of the population being sampled. The 200 plot recommendation will generally achieve a standard error of about 12% for estimates of the number of trees in a city-wide sample. This is based on the research in this publication, https://www.itreetools.org/documents/14 ... 20size.pdf. You are correct that natural forests generally have less heterogeneity and typically require fewer plots to obtain similar sampling precision or standard errors. If you are interested in trying to establish an optimal sample size (i.e. the best number of plots to reach your desired standard error) you may want to check out this forum post, viewtopic.php?f=19&t=1210.
Thanks for your question.
-Jason