Shawn Fraver

Date of Award


Level of Access Assigned by Author

Campus-Only Dissertation

Degree Name

Doctor of Philosophy (PhD)


Forest Resources


Alan S. White

Second Committee Member

Charles V.H. Cogbill

Third Committee Member

Malcolm L. Hunter


I linked methods of dendroecology with spatial analyses to investigate the patterns of natural disturbance in a ca. 2000-ha old-growth landscape of northern Maine. I randomly placed 38 plots among five community types throughout the landscape and reconstructed a history of disturbance (a chronology) for each. I found no evidence of stand replacing disturbance during the period covered by this investigation (back to the mid 1700s). The disturbance rates were generally low (mean of 9.6% canopy destruction per decade), yet fluctuated markedly through time. Minor peaks in disturbance were synchronized among plots, albeit weakly, suggesting landscape-wide pulses of tree recruitment, canopy openness, and deadwood abundance. However, I found no evidence of spatial patterning of disturbance: no correlation between disturbance patterns and the physical distance between plots, and no evidence of disturbance patch structure. Matrix correlation indicated that species composition best explained the variability in temporal patterns of disturbance. Neither landform nor aspect influenced rates or patterns or disturbance. The disturbance dynamic appears to have two components: pulses of moderate severity disturbances caused by host-specific disturbance agents (primarily the spruce budworm) interposed upon a background of scattered small-scale canopy gaps. The relatively low rates of canopy disturbance, coupled with the abundance of advance regeneration, have maintained canopy dominance by shade-tolerant species. Peaks of low- to moderate-severity disturbance may admit mid-tolerant tree species; however, they are insufficient to admit shade-intolerant species. I developed an alternative method of detecting growth releases fiom tree-ring series, based on absolute increases in radial growth. The method has convenient mathematical properties that overcome several shortcomings in the commonly used percent-increase method. By intensively sampling and mapping three red spruce (Picea rubens) plots, I determined the locations and sizes of former canopy gaps back to 1920. Gaps were generally small (median 25.3 m2). Red spruce trees clearly benefit fiom these gaps: 80.6% of trees showed one or more releases before canopy accession. Several measures of canopy structure provided strong supporting evidence for historical disturbance patterns revealed by dendroecological methods.