We review the relative merits and field procedures for four basic plot
designs to handle forest inventory plots that straddle two or more co
nditions, given that subplots will not be moved. A cluster design is r
ecommended that combines fixed-area subplots and variable-radius plot
(VRP) sampling. Each subplot in a cluster consists of a large fixed-ar
ea subplot for estimating area and tallying large trees, a microplot (
small fixed-area subplot) for tallying small trees, and a variable-rad
ius plot for tallying intermediate size trees. Subplots are assigned a
condition class based on a classification algorithm. For subplots tha
t straddle conditions, boundaries are mapped for area estimation, and
the condition class of each tree is recorded. For some straddler subpl
ots, there may be insufficient trees for specific conditions to allow
unambiguous classification by computer algorithm. We discuss several p
ossible solutions for this problem.