Document: AND-3-35-46

Sampling structure and diversity of old-growth forests in the Pacific Northwest: Effect of plot size.

GRAY, A.N.*

USDA Forest Service Pacific Northwest Research Station, Portland OR 97205 USA ¹

Abstract:
Renewed interest in the regional distribution and trends of forest habitat structure and diversity have placed demands on forest inventories for accurate stand-level data. A primary need in the coastal Pacific Northwest is information on the extent and rate of development of late-successional forest structure. The objective of this study was to devise a sampling scheme, within a national cluster-plot design, able to efficiently determine stand-level tree attributes. A simulation approach was taken, using stem maps from 13 permanent forest plots which were at least 1 ha in size and had been sampled for 15-20 yrs. Simulated subplots, ranging from 2 to 25 m in radius, were randomly located within stand boundaries to select mapped trees. Data from sets of four subplots were grouped into virtual plots and the sampling repeated thirty times per subplot radius per stand. Calculated sample attributes were: Density of live trees by diameter classes, annual tree mortality, and tree species diversity. Sample error analysis compared the percent deviation between sample data and full-stand values by subplot size. For live tree density in most size classes, subplots of at least 14 m were required to reduce errors below 30% of true density 60% of the time. For mortality of large trees, subplot sizes 20 m were needed to achieve errors within 80%. Subplots of 20 m in radius were also necessary to accurately portray tree species richness, which was concentrated in the smaller tree sizes. The relatively large plot sizes required for accurate sampling of stand attributes most likely reflect the patchy structure and composition of old-growth forest stands. Nevertheless, a nested plot design sampling large trees and snags on large plots can be used to efficiently characterize stand structure. Reliable characterization of tree diversity, however, should probably augment standard plot sampling with stand-level species lists.

Keywords:

This abstract is being presented at: 10:30 AM in session:
STATISTICAL ECOLOGY