Effectiveness of nearest-neighbor data adjustment in a clonal test of redwood

Tree-to-tree environmental correlations were estimated for various traits in a clonal experiment with coast redwood (Sequoia sempervirens (D. DON) ENDL.). A nearest-neighbor adjustment was performed to reduce the effects on individual-tree data of environmental gradients and patchiness within the test site. The size of the environmental correlation proved reliable as a predictor of the usefulness of neighbor adjustment. Some results of the neighbor adjustments were: (a) clonal means and rankings for metric traits were changed, in some cases substantially so; (b) for traits with neighbor-tree correlations greater than 0.15, estimates of within-clone components of variance were substantially decreased; (c) estimates of among-clone components of variance were moderately but consistently increased; (d) broad-sense heritabilities and predicted genetic gains were thus increased. Estimates of among-stand components of variance were little changed.