What is the most efficient and effective method for long-term monitoring of alpine tundra vegetation?

Nondestructive estimations of plant community characteristics are essential to vegetation monitoring programs. However, there is no universally accepted method for this purpose in the Arctic, partly because not all programs share the same logistical constraints and monitoring goals. Our aim was to determine the most efficient and effective method for long-term monitoring of alpine tundra vegetation. To achieve this, we established 12 vegetation-monitoring plots on a south-facing slope in the alpine tundra of southern Yukon Territory, Canada. Four observers assessed these plots for vascular plant species abundance employing three methods: visual cover (VC) and subplot frequency (SF) estimation and modified point-intercept (PI) (includes rare species present but not intersected by a pin). SF performed best in terms of time required per plot and sensitivity to variations in species richness. All methods were similarly poor at estimating relative abundance for rare species, but PI and VC were substantially better at high abundances. Differences among methods were larger than among observers. Our results suggest that SF is best when the monitoring focus is on rare species or species richness across extensive areas. However when the focus is on monitoring changes in relative abundance of common species, VC or PI should be preferred.