Simulating changes in forest structure and light availability in response to different thinning approaches

Thinning dense, second-growth stands to accelerate development of old-growth forest conditions is receiving increasing attention in the Pacific Northwest.  However, the diverse goals implicit in these manipulations are not always mutually compatible (e.g., maximizing growth of larger trees and increasing spatial heterogeneity of structure or understory light).  The ability to predict the effects of manipulations on forest structure and resource availability may be useful in designing treatments with specific objectives.  We used stem maps from 1- to 2-ha plots in second-growth stands in western Washington and simulated five thinning treatments.  We then used Morisita’s index and the light-interception model tRAYci to predict changes in the spatial structure of the forest and in light conditions 1 m above the forest floor.  Each treatment removed 30% of the original basal area, but differed in how trees were removed:  (1) neutral thin removed trees at random; (2) thin from below removed the smallest trees; (3) random ecological thin protected the largest trees constituting 40% of the basal area and removed half of the remaining trees randomly; (4) structured ecological thin also protected the largest trees, but removed half of the remaining trees in clumps; and (5) gaps removed all trees in circles ~1 tree height in diameter.  Neutral thin, thin from below and random ecological thin produced generally similar understory light distributions, but thin from below produced a much more regular spatial structure than the other two treatments.  Structured ecological thin produced a more variable light distribution and a more clumped spatial structure than the other thinning treatments.  Gaps produced the broadest light distribution.  These simulations provide insights into ways in which different types of treatments produce different post-harvest conditions and potentially different ecological outcomes.