A boundary-line approach to dendroecological release criteria
Open Access
- Author:
- Black, Bryan Andrew
- Graduate Program:
- Forest Resources
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- September 18, 2003
- Committee Members:
- Marc David Abrams, Committee Chair/Co-Chair
Kim C Steiner, Committee Member
Roger Tai Koide, Committee Member
David Eissenstat, Committee Member - Keywords:
- dendrochronology
historical ecology
forest dynamics
disturbance ecology
release criteria
eastern hemlock - Abstract:
- Identifying releases from suppression represents one of the most fundamental dendroecological procedures for quantifying forest disturbance histories. In tree-ring series, releases are typically defined as pulses in percent-growth change that exceed a minimum threshold. Past methodologies have applied fixed growth change thresholds that may overly generalize the ability of a tree to respond to a disturbance event. This study presents a more precise method that scales each release by its maximum release potential as defined by prior growth rates, using an old growth hemlock (Tsuga canadensis L.) forest in central Pennsylvania as an example. On average, tree age, diameter, and crown class appear to influence the magnitude of release response. However, differences in release response between age, diameter, and canopy classes are largely an artifact of the more fundamental relationship between release response and growth immediately prior to release. Specifically, maximum percent-growth change declines at a steep, negative exponential rate as prior growth levels increase. Slow-growing hemlocks can exceed 800% growth increases while fast-growing hemlocks do not exceed growth increases of twenty percent. This negative exponential threshold, or boundary line, represents the maximum percent-growth change that is physiologically possible at a given level of prior growth. Thus, release potential is relative to prior growth rate. The relationship between prior growth and release potential is species-specific and is demonstrated in 24 additional hemlock stands. New release criteria based on prior growth are proposed in which moderate and major releases are those falling within 20-49.9%, and 50-100% of the boundary line, respectively. Effects of short, moderate, and long-term climate events are reduced by the percent growth change calculation and the high growth-change threshold applied to slow-growing trees. Overall, this approach provides a more theoretically sound and flexible approach to developing disturbance histories. Prior growth is also shown to strongly influence release response in species representing a wide diversity of habitats in North America. Species included in this study are white pine (Pinus strobus), chestnut oak (Quercus prinus), white oak (Quercus alba), shortleaf pine (Pinus echinata), white spruce (Picea glauca), black spruce (Picea mariana), bur oak (Quercus macrocarpa), post oak (Quercus stellata), ponderosa pine (Pinus ponderosa), and Douglas-fir (Pseudotsuga menziesii). Prior-growth boundary lines are developed for each species, and in all cases the boundary line follows a negative exponential function. Size and diameter interactions with release response are then explored and reveal that diameter or age limits may be necessary for many species. In general, shade-intolerant species, which grow where competition is low due to fire or other edaphic factors, tend to lose release potential the fastest as age and size increase. Tolerant species retain full release response for nearly all of their lives. Furthermore, boundary line differences among northeastern species reflect levels of understory tolerance in that shade tolerant individuals have a relatively high release potential under slow-growing conditions while intolerant individuals have a relatively high release potential under fast-growing conditions. Thus prior growth not only can be applied to enhance release criteria, but also to reveal differences among tolerant and intolerant species. A bootstrapping test is designed to estimate the minimum sample size necessary to develop a species-specific boundary line. The results somewhat consistently indicate a minimum sample size of 50,000 growth increments and an ideal sample size of 150,000 growth increments to develop a prior-growth boundary line. Next, the boundary-line release criteria are validated using tree-ring data obtained from a chestnut oak stand with a known thinning history. The largest crown releases (87.5 – 100% of crown exposed) approach the boundary line while smaller crown releases occupy increasingly lower subsets of the boundary line maximum value. This validates that the boundary line represents maximum release potential and that the degree of release is proportional to distance from the prior-growth boundary line. Release responses of nineteen white oak trees that survived a catastrophic tornado also approach the white oak boundary line, but less consistently. Failure to reach the boundary line is correlated with small difference in radius, which may be due to the exposure of these trees to the high winds. If these large trees have muted or delayed release responses, the estimated date of a catastrophic disturbance may be much later than the true date of the event. Overall, these examples illustrate that incorporation of prior growth into release criteria could greatly improve the ability to detect disturbances and estimate the magnitude of each disturbance event. Thus these new criteria represent a significant improvement over pre-existing methodologies and will have broad applications to many other species in a wide variety of forest types.