STAND AND LANDSCAPE-LEVEL DISTURBANCE DYNAMICS IN OLD-GROWTH FORESTS IN WESTERN MASSACHUSETTS
Natural disturbances strongly influence the dynamics and developmental patterns of forest ecosystems; however, relatively little is known about the historic patterns of natural disturbance for many portions of eastern North America, such as southern New England, where human disturbance has predominated for centuries. For these regions, much of our understanding of natural disturbance dynamics comes from studies of younger second-growth forests or isolated old-growth stands, thus limiting the temporal and spatial resolution of our knowledge of dynamics in these regions. To address these limitations, we analyzed dendroecological data from the 18 largest remaining old-growth stands in western Massachusetts, ranging in proximity from 1–60 km apart, in an effort to characterize the historic stand and landscape-level patterns of natural disturbance. Our results indicate that disturbance regimes for these systems were dominated by relatively frequent, low-intensity disturbances (average 5.0% \textpm 0.2% canopy area disturbed per decade) operating somewhat randomly on the landscape. Across the study areas, most decadal disturbances (86.2%) involved <10% canopy loss. There was no evidence of stand-replacing disturbances during the period examined (1700–1989), and the maximum canopy area disturbed in any given decade was 26.3%. Nonmetric multidimensional scaling demonstrated that several forests shared similar disturbance histories despite being separated by >50 km. Comparisons of these patterns with model simulations of past hurricane events and historical documents suggest that broadscale disturbances, such as hurricanes and ice storms, resulted in common disturbance peaks and subsequent recruitment peaks at spatially disparate areas in the 1790s, 1870s, 1900s, and 1920s. Conversely, the lack of synchrony in proximate areas during these events highlights the patchy nature of these disturbances on the landscape. Compositional and physiographic factors influenced disturbance patterns, as stands located on northwest-facing slopes or containing significant Picea rubens components in the forest overstory experienced the highest levels of disturbance. Our results highlight the utility of incorporating dendroecological reconstructions across numerous old-growth stands to interpret the historic stand and landscape-level disturbance dynamics in areas devoid of large, contiguous old-growth landscapes.