Increases in tree mortality in many North American forests during recent decades have raised concerns about resilience in the face of a changing climate.  Thinning treatments have been suggested as a strategy for reducing mortality rates, but responses to thinning likely vary across treatments that produce different distributions of tree sizes and densities.  We analyzed 46 years of mortality data from a thinning experiment in red pine (Pinus resinosa) stands that tested various combinations of thinning methods and intensities to understand how trends in climate and stand structure contribute to differences in mortality among thinning treatments.  Treatments included thinning from above (removal of trees in dominant and codominant crown positions), thinning from below (removal of trees in suppressed and intermediate positions), and a combination treatment (removal of trees throughout the canopy profile) applied across several levels of residual basal area (7, 14, 21, 28, and 35 m² ha^-1) and unharvested control stands.

Results/Conclusions

Mortality increased over time in unharvested stands, along with increases in mean annual and growing season temperatures, mean tree sizes, and stand basal area, while stem densities decreased.  Mortality generally remained lower in stands that were thinned from above or below, except in treatments that combined thinning from above with low basal areas, which had very high mortality in the 5-10 years following initial thinning.  Stands that were thinned from below had low mortality throughout the study, regardless of basal area.  Over time, stem densities became higher and mean tree sizes smaller in stands that were thinned from above compared to stands that were thinned from below.  In general, trees in thinning treatments became larger and stem densities lower than trees in unharvested stands.  Individual tree mortality models indicated that trees with higher relative growth rates, larger relative diameters, and more open neighborhood environments had much lower risk of death than smaller, slower-growing trees in denser environments.  This could explain why mortality was so low in stands that were thinned from below, which favored retention of larger, widely-spaced, vigorously-growing trees.  Tree size – density relationships suggest regular density-dependent interactions likely contributed to increased mortality in unharvested stands, so we cannot definitively link mortality trends to rising temperatures.  Results from this long-term study do demonstrate, however, that thinning from above to moderate or high basal areas and thinning from below across a wide range of basal areas effectively reduced mortality, even during a period of warming.