Adaptation and mitigation capacity of wildland forests in the northeastern United States

Abstract

Managing forests to mitigate climate change and increase their capacity to adapt to future climate-related disturbances and conditions typically involves protecting and enhancing forest carbon stocks and sequestration capacity while promoting structural diversity. While the focus has been on comparing active management approaches for meeting these objectives, there are few empirical assessments of passive management. Here we used quasi-experimental methods to compare carbon and structural complexity within “wildlands,” where harvesting and other land uses are prohibited, to environmentally comparable forests without protection from timber harvesting. Using USDA Forest Inventory and Analysis (FIA) plots from the Adirondack-New England region of the Northeastern U.S., we compared aboveground carbon, total forest basal area increment (our proxy for carbon sequestration), and six forest-level structural variables in forests. To help explain observed differences, we examined (1) the recent history of harvesting within unprotected forests, (2) stand age in wildland and unprotected forests, and (3) the carbon and structural attributes of protected and unprotected plots at the initiation of wildlands protection. Aboveground carbon was 20% higher in wildlands overall (P < 0.0001), with differences greatest in wildlands of New York (+32%; P = 0.0001) and in Maine (+34%; P = 0.01) where recent harvesting intensity and differences in stand age between protection categories were highest. Basal area increment did not differ between protected areas at the regional and sub-regional scale, but was 37% higher in wildlands (P = 0.03) than in recently harvested areas. Structural complexity was generally higher in wildlands, with four structural variables – large live (>60 cm DBH) and large dead (>45 cm DBH) tree density, maximum tree height, and diversity of diameter size classes) – greater in wildlands than in unprotected forests. Two variables (adult tree species richness and standard deviation of tree height) did not differ between protection categories. Both carbon and structural differences were amplified by recent harvesting in unprotected plots. For the subset of plots that allowed for comparison, wildlands did not differ in carbon and structural attributes from unprotected plots at the onset of wildlands protection, suggesting that subsequent management rather than initial differences was the driver of carbon and structural differences between protection categories. Our results highlight the adaptation and mitigation benefits of allowing natural processes to predominate in strictly protected areas.