Responses of planted and naturally regenerating conifers in a variable-retention experiment in the Pacific Northwest

Background/Question/Methods

Variable-retention harvest is a key component of ecosystem-based forest management worldwide.  It involves the retention of structural elements (live trees, snags, or logs) that are intended to emulate the outcomes of natural disturbances.  Two elements of residual forest structure—the level (amount) and spatial pattern of retained trees—are likely to play important roles in the post-harvest regeneration of forests, but long-term studies of regeneration responses are rare.  Partial retention of overstory trees is hypothesized to have positive effects by ameliorating understory microclimate and by enhancing seed availability.  Conversely, retained trees may inhibit regeneration, particularly for shade-intolerant species.   We present the results of a regional-scale experiment in variable retention of mature coniferous forests of the Pacific Northwest—the DEMO study.  Replicated at six locations in western Oregon and Washington, it offers a rare opportunity to study regeneration responses to both the level of retention (40 vs. 15% of original basal area) and its spatial distribution (dispersed vs. aggregated) across a wide range of forest types and physical environments.  We assess the performance of planted seedlings (mortality and height growth of three species) and the survival and recruitment of natural regeneration of two height classes (<10 cm and > 10 cm, but < 5 cm dbh) including both early- and late-seral species.

Results/Conclusions

Both level and pattern of retention had significant effects on planted seedling performance.   Early mortality (1-2 yr) was generally reduced in dispersed treatments (40D or 40D and 15D) compared to the cleared areas of aggregated treatments (40A-c and 15A-c).  Height growth, however, was suppressed at 40% retention (either 40D or 40D and 40 A-c), but effects were not manifested until later in the study (5-12 yr).  Natural regeneration was also sensitive to level and pattern of retention.  Regeneration densities were higher in dispersed treatments.  Total density was highest in 15D (mean of 4650/ha) as was the density of Pseudotsuga menziesii, the early-seral dominant (mean of 3183/ha).  However, 40D had the highest density of late-seral species (mean of 2094/ha), significantly greater than the density in 40A-c (mean of 972/ha).  Our long-term studies of regeneration outcomes indicate that level and pattern of overstory retention can have substantial effects on survival and growth of planted trees and on the density and composition of natural regeneration.  Both sets of outcomes have important implications for the longer-term development of forest composition and structure.