Tuesday, August 7, 2007 - 4:00 PM

COS 43-9: Informing restoration thinning prescriptions with spatial point process models

Andrew J. Larson, University of Washington

Thinning is a widely used silvicultural tool for accelerating the development of late-successional and old-growth structure in previously harvested forests.  Virtually all restoration thinnings, however, result in a very regular spatial pattern—hard core inhibition—of residual trees at local scales (<3 m).  Residual trees left following restoration thinning are intended to be the initial cohort of overstory trees as the restored forest develops old-growth structure.  The spatial pattern of overstory trees in natural old-growth Abies amabilis forests differs from a hard core inhibition pattern (P<0.05), suggesting that current restoration thinning prescriptions commonly used in this forest type homogenize spatial patterns at small scales relative to old-growth forests.

Spatial patterns of overstory trees in old-growth Abies amabilis forests were modeled with a soft core inhibition spatial point process model.  Inhibition strength was represented as a function of tree size and decayed linearly with distance.  Model parameterizations were found such that patterns produced by the model were not different from the observed pattern of overstory trees in two ~300 year old stands.  Interestingly, the pattern of overstory trees in a ~600 year old stand could not be reproduced.

Spatial point process models are collections of rules for assigning locations to points on a plane.  Similarly, thinning prescriptions are rules for designating trees to be left following thinning, and, either implicitly or explicitly, where to leave them.  Restoration thinning prescriptions may be improved by formulating them as approximations of parameterized spatial point process models of overstory trees in natural old-growth stands.