|Title||Estimates of downed woody debris decay class transitions for forests across the eastern United States|
|Publication Type||Journal Article|
|Year of Publication||2013|
|Authors||Russell, Matthew B., Woodall Christopher W., Fraver Shawn, and D’Amato Anthony|
|Pagination||22 - 31|
|Keywords||bioenergy, biomass, carbon flux, coarse woody debris, forest inventory|
Large-scale inventories of downed woody debris (DWD; downed dead wood of a minimum size) often record decay status by assigning pieces to classes of decay according to their visual/structural attributes (e.g., presence of branches, log shape, and texture and color of wood). DWD decay classes are not only essential for estimating current DWD biomass and carbon stocks, but may also facilitate the prediction of future DWD attributes. Estimating temporal transitions between decay classes may provide a mechanism for projecting DWD attributes in forest ecosystems. To date, modeling decay class transitions for individual DWD pieces has not been fully explored in this context. The goal of this study was to use a repeated DWD inventory across the eastern US to estimate decay class transitions to inform DWD dynamics across this broad geographic region. Using matched and non-matched DWD from the repeated inventory, ordinal regression techniques were used to estimate the five-year probability of a DWD piece remaining in the same decay class or moving into more advanced decay classes. Models indicated that these transitions were largely related to DWD piece length and climatic regime, as transitions occurred more slowly for longer DWD pieces located in regions with a low number of degree days (a climatic variable serving as a proxy for decomposition potential). Cumulative link mixed models allowed the estimation of forest type-specific effects (i.e., random effects) on the DWD transition process. Hardwood species transitioned into subsequent decay classes more rapidly than softwoods. Model assessments indicated that the correct decay class observed after five years was correctly predicted for approximately 50-70% of observations, but was dependent on forest type and initial decay class. Results differed depending on the models under examination. For example, using the matched data, the average number of classes moved per five years was 1.28 +/- 0.07 (mean +/- SE) classes for decay class 1 logs found in spruce-fir forests, however, using the matched plus non-matched data, the average number of classes moved per five years was 3.51 +/- 0.19 for these same logs. These two model sets (matched and matched plus non-matched DWD pieces) may denote upper and lower bounds for DWD decay class transition rates. Analyses presented herein provide an initial assessment of DWD decay across eastern US forests and thus provide quantitative tools that apply to emerging bioenergy questions and associated DWD dynamics research. Developed models, coupled with traditional forest productivity simulation tools, may be used in the future to determine accurate estimates of future forest C stocks.