A METHOD FOR MONITORING LONG-TERM POPULATION TRENDS - AN EXAMPLE USING RARE ARCTIC-ALPINE PLANTS

Populations of arctic-alpine plants at the southern periphery of their range should respond rapidly to projected global warming. We establis hed a study to monitor the density and reproductive effort of six such species in tundra of Glacier National Park, Montana to help project t he fate of these species in the center of their range. For many specie s, detecting long-term population trends is confounded by short-term v ariation. Our study design employs temporal resampling of permanent pl ots on multiple sites and a repeated-measures model that accommodates the effects of high frequency variation and allows assessment of the s ignificance of long-term trends. Statistical analysis compares site-sp ecific estimates of annual mean density between two time periods and u ses between-plot, within-site, within-year variation to estimate error . Power calculations using data from the 3-yr baseline period indicate that measurements of fecundity will be less sensitive for detecting l ong-term trends than measurements of plant density. Furthermore, our r esults suggest that perennial species of closed-turf communities may b e better bioassays of long-term change than annuals or species of open , ephemeral microsites.