Forest Science

VanderSchaaf, Curtis L.

doi: 10.1093/forestscience/56.4.327pmid: N/A

Understanding self-thinning patterns of a species helps resource managers make decisions about proper planting densities and timings of thinnings. It is generally assumed that size–density trajectories of even-aged, monospecific, self-thinning stands consist of three phases: first, a curved approach to a linear portion that represents moderate rates of intraspecific mortality; second, a linear portion where tree density/ha (N) is at its maximum for a particular quadratic mean diameter (Dq), often termed the maximum size–density relationship (MSDR) dynamic thinning line; and third, a divergence from the linear portion. In this article, segmented regression was used to determine what observations are within various phases of self-thinning for a loblolly pine (Pinus taeda L.) planting density trial located in central Mississippi. Those observations estimated to be within the MSDR dynamic thinning line phase of individual plots were combined to estimate a MSDR species boundary line slope (−1.640) using a linear mixed-effects model approach. Based on the N and Dq, where the MSDR dynamic thinning line phase was estimated to begin using segmented regression analyses, and the estimated MSDR species boundary line slope of −1.640, planting density-specific maximum values of Reineke's stand density index were predicted.

Keim, Richard F.; Dean, Thomas J.; Chambers, Jim L.; Conner, William H.

doi: 10.1093/forestscience/56.4.336pmid: N/A

Stand density indices (SDIs) are surrogate measures of site occupancy useful for developing quantitative management tools for forest types that have not been extensively studied. For example, there has been little research of growth-density relationships for baldcypress (Taxodium distichum [L.] L.C. Rich.). We used published data, forest inventory data, and a 25-year-old thinning study to identify maximum densities and density of incipient self-thinning in baldcypress. Results suggest that a maximum Reineke's SDI is approximately 1,200 (SI units) in baldcypress, which agrees with practical experience and the theoretical limit derived from observed correlations between a species wood density and maximum SDI. Thinning a baldcypress stand to ≥58% of maximum SDI did not affect tree-level or stand-level growth, but thinning to 40% of maximum SDI increased diameter increment and net volume increment (at α = 0.07) in the first 25 years after thinning. These results are consistent with expectations based on stand density-growth relationships for other species, which supports the general utility of management tools based on stand density concepts.

Polyakov, Maksym; Wear, David N.; Huggett, Robert N.

doi: 10.1093/forestscience/56.4.344pmid: N/A

Timber supply has traditionally been modeled using aggregate data, whereas individual harvest choices have been shown to be sensitive to the vintage and condition of forest capital stocks. In this article, we build aggregate supply models for four roundwood products in a seven-state region of the US South directly from stand-level harvest choice models applied to detailed forest inventories. These models allow for a more precise accounting of the biological and economic underpinnings of supply and support forecasting of changes in forest inventories with a high degree of detail. Estimation results support use of the approach. The elasticities of softwood and hardwood sawtimber supply, 0.34 and 0.31, respectively, are consistent with the elasticities reported by previous studies. The elasticities of softwood and hardwood pulpwood supply (respectively, 0.062 and 0.025) are much lower than previous studies found for pulpwood supply, and cross-price elasticities indicate a dominant influence of sawtimber markets on pulpwood supply. Results generally indicate complementarity between sawtimber and pulpwood supply in the short run. This approach can provide a means of predicting the supply consequences of exogenous factors that could alter forest inventories, e.g., climate change and invasive species, and support regular updating of supply models as new inventory data are recorded.

Lyons, Kevin; Xing, Li; Nelson, John D.

doi: 10.1093/forestscience/56.4.356pmid: N/A

This article applies a linear mixed-effects model (LME) to an unpublished helicopter logging productivity data set. The data were clustered in units, and the silviculture treatments varied between units, where units are cutblocks. When the covariance matrix was considered, it was found that the interunit variance was significant and the intraunit variance was heterogeneous. The significant interunit variance results in significant intraunit correlation, and this indicates the need to use LME to analyze these data. When turn time was considered, unit 4, which had the lowest level of group retention, had a conditional predicted value that was significantly less than the marginal predicted value. When turn weight was considered, unit 2, which had small patch cuts, had a conditional predicted value that was significantly less than the marginal predicted value, and unit 4 had a conditional predicted value that was significantly higher than the marginal predicted value. When productivity was considered, the conditional predicted value for unit 4 was significantly greater than the marginal predicted value, and unit 2 was significantly less than the marginal predicted value. In this data set it is interesting to note that productivity is independent of the number of logs in a turn and that turn weight was relatively insensitive to the available explanatory variables.

Stennes, Brad K.; Niquidet, Kurt; van Kooten, G. Cornelis

doi: 10.1093/forestscience/56.4.366pmid: N/A

Energy has been produced from woody biomass in British Columbia for many decades, primarily within the pulp and paper sector, using residual streams from timber processing to create heat and electricity for on-site use. More recently, there has been some stand-alone electricity production and an increase in the capacity to produce wood pellets, both using “waste” from the sawmill sector. Hence, most of the low-cost feedstock sources associated with traditional timber processing are now fully employed. Although previous studies modeled bioenergy production in isolation, we used a fiber allocation and transportation model of the British Columbia forest sector with 24 regions to demonstrate that it is necessary to consider the interaction between use of woody feedstock for pellet production and electricity generation and its traditional uses (e.g., production of pulp, oriented-strandboard, and others). We find that, despite the availability of large areas of standing mountain pine beetle-killed timber, this wood does not enter the energy mix in a dedicated salvage timber harvest to energy system. Further expansion of biofeedstock for energy is met by a combination of woody debris collected at harvesting sites and/or bidding away of fiber from existing users.

Podlaski, Rafał

doi: 10.1093/forestscience/56.4.379pmid: N/A

The objectives of this study were to investigate the suitability of two-component Weibull and gamma mixtures to model the dbh distribution of a mixed-species, two-age cohort stands and for age cohort determination and to compare several methods to choose initial parameter values for maximum likelihood estimation of mixture models. Investigations were carried out in near-natural, fir (Abies alba Mill.)–beech (Fagus sylvatica L.), two-age cohort stands in the Świętokrzyski National Park (Central Poland), where unusually high mortality of fir followed by its recovery and revitalization has been observed. The age cohort YG1 is composed of trees from 60 to ∼150 years of breast height age, and the age cohort YG2 is composed of trees less than 60 years of breast height age. The empirical distributions for the stands in this study were equally well fit by both the mixture Weibull and gamma models. It has been assumed that the estimated values, the weights (fractions), the means, and the standard deviations of two-component mixture models, are the predicted values of dbh statistics of age cohorts. The mean absolute relative errors used to evaluate this assumption were least for age cohort YG2 (from 14.8 to 29.6%) and largest for age cohort YG1 (from 17.7 to 45.0%). The dbh component 1 of mixture models can be identified in the stands investigated with age cohort YG2 and to a lesser degree the dbh component 2 with age cohort YG1. The multistart method for choosing initial values for the numerical procedure (a combination of the expectation-maximization algorithm with the Newton-type method) was best but also the most labor-intensive. The optimal way to estimate parameters in two-component mixtures with the Weibull or the gamma distributions is to apply min/max and 0.5/1.5/mean methods and, additionally, but only if necessary, a multistart method.

Marcot, Bruce G.; Ohmann, Janet L.; Mellen-McLean, Kim; Waddell, Karen L.

doi: 10.1093/forestscience/56.4.391pmid: N/A

We used novel methods for combining information from wildlife and vegetation field studies to develop guidelines for managing dead wood for wildlife and biodiversity. The DecAID Decayed Wood Advisor presents data on wildlife use of standing and down dead trees (snags and down wood) and summaries of regional vegetation plot data depicting dead wood conditions, for forests across the Pacific Northwest United States. We combined data on wildlife use by snag diameter and density and by down wood diameter and cover, across studies, using parametric techniques of meta-analysis. We calculated tolerance intervals, which represent the percentage of each species' population that uses particular sizes or amounts of snags and down wood, and rank-ordered the species into cumulative species curves. We combined data on snags and down wood from > 16,000 field plots from three regional forest inventories and calculated distribution-free tolerance intervals compatible with those compiled for wildlife to facilitate integrated analysis. We illustrate our methods using an example for one vegetation condition. The statistical summaries in DecAID use a probabilistic approach, which works well in a risk analysis and management framework, rather than a deterministic approach. Our methods may prove useful to others faced with similar problems of combining information across studies in other regions or for other data types.

Duane, Maureen V.; Cohen, Warren B.; Campbell, John L.; Hudiburg, Tara; Turner, David P.; Weyermann, Dale L.

doi: 10.1093/forestscience/56.4.405pmid: N/A

Empirical models relating forest attributes to remotely sensed metrics are widespread in the literature and underpin many of our efforts to map forest structure across complex landscapes. In this study we compared empirical models relating Landsat reflectance to forest age across Oregon using two alternate sets of ground data: one from a large (n ∼ 1500) systematic forest inventory and another from a smaller set of plots (n < 50) deliberately selected to represent pure conditions along predefined structural gradients. Models built with the smaller set of targeted ground data resulted in lower plot-level mapping error (root mean square error) and higher apparent explanatory power (R2) than those built with the larger, more widely distributed inventory data. However, in two of the three ecoregions considered, predictions derived from models built with the smaller ground data set displayed a bias relative to those built with the larger but noisier inventory data. A modeling exercise, wherein mapped forest age was translated into carbon, demonstrated how nonlinear ecological models can magnify these prediction biases over landscapes. From this study, it is clear that for mapping purposes, inventory data are superior to project-specific data sets if those data sets are not representative of the full region over which mapping is to be done.

Loehle, Craig

doi: 10.1093/forestscience/56.4.417pmid: N/A

Detectability issues create uncertainty in field surveys of populations. Methods to overcome this problem include mark-recapture methods and double sampling. Partial double sampling involves estimating detectability, [inline-graphic not available: see fulltext], and using this to correct the estimate in areas sampled only once. Results of this study indicate that, if population density is not uniform across sample units, there is a tradeoff for a given sampling effort, with lower population estimation error for double sampling only a portion of the area when spatial variance and/or detectability are lower but lower error for partial double sampling when spatial variance and/or detectability are higher. Thus, whereas standard double sampling increases precision, it does not give adequate accuracy in the face of spatial variation unless the entire area can be surveyed twice, which is more expensive.