Two basic monitoring designs are available for detecting changes in pl
ant density for monitoring studies: temporary and permanent plot. Ther
e are advantages to both methods. We used data from 3-year monitoring
studies for 12 species to compare the designs with respect to sample s
ize requirements. For each of the designs, statistical models of plant
density and hypothesis tests were formulated. Monitoring data were us
ed to compute the approximate minimum number of plots required to main
tain a set level of power for both designs. The temporary plot designs
required at least twice as many plots as the permanent plot designs f
or 10 of 12 species. However, the advantage of permanent plot designs
was greatest for long-lived perennials and least for annuals and short
-lived perennials. We discuss the trade-offs involved when choosing be
tween these two sampling methods.