Concern about global environmental change and associated uncertainty has given rise to greater emphasis on the importance of forest ecosystem resilience and building adaptive capacity through management. Greater diversity is generally believed to contribute to increased resilience, but that relationship is nuanced and varies widely across ecosystems and scales. Thus, knowing how to implement objectives for increasing or maintaining resilience remains a challenge.
We examined the impact of time and management on functional composition, functional diversity, and annual trait-based growth in three northern forests distributed across the northeastern United States. We accomplished this using a combination of historic inventories and dendrochronological techniques. Two harvest treatments (typical for each forest type and region) and an untreated control were replicated at least three times at each of three sites: the Argonne, Bartlett, and Penobscot Experimental Forests (EF) located in Wisconsin, New Hampshire, and Maine, USA, respectively. Increment cores were collected from all trees with diameter at breast height > 10 cm in each 0.04 ha plot. Trait means (leaf mass per area, specific gravity, maximum height, seed mass, drought tolerance, shade tolerance, and flood tolerance), standard functional diversity measures, and relative density were calculated annually 1984-2009 based on biomass increment for all trees in each plot, estimated using species-specific allometric equations.
Results/Conclusions
Harvest treatments influenced functional composition and functional diversity across sites with varying trends. At the Argonne EF, the dispersion of effect traits was greater for controls and the crop tree release treatment than for selection based on 2008 standing biomass; no differences in 2008 functional diversity were observed in response to treatment at the Penobscot EF or Bartlett EF. Change in functional dispersion for effect and response traits (1984-2008) was greatest for the shelterwood treatment at the Penobscot EF and differed from other treatments.
At the Argonne EF and Bartlett EF, annual trait-based growth in mature, relatively undisturbed forest stands remained stable with no significant effects of time and limited effects of relative density observed. At the Penobscot EF, however, the dispersion and divergence of effect traits as expressed through annual growth declined over the course of 25 years. Significant, positive correlations (p<0.001) were observed between annual stand-wide biomass increment, an approximation of productivity, and annual increases in the dispersion (Argonne EF and Penobscot EF) and divergence (all three sites) of response traits.