Trees accumulate biomass by sequestrating atmospheric carbon and allocate it to different tree components. A biomass component ratio is the ratio of biomass in a tree component to total tree biomass. ...Modeling the ratios for Zelkova serrata, an important native reforestation tree species in Taiwan, helps in understanding its biomass allocation strategy to design effective silvicultural treatments. In this study, we applied Component Ratio Method (CRM) to relate biomass component ratios of main stem, large branch, twig, and foliage to tree attributes of Z. serrata from a 9-year-old plantation. Nonlinear and linear CRM models were fitted with Seemingly Unrelated Regression to account for model correlations. Linear CRM models with dbh as the predictor had the best fit with model correlations as high as 80%. About 46% and 40% of total tree biomass was allocated to main stem and large branch, respectively. However, main stem biomass decreased by 1.9% with every 1-cm increase in dbh, but large branch biomass increased by 2.2% instead. Results suggest that dominant Z. serrata trees tend to branch and fork, while smaller trees invest in larger main stem. An early pruning treatment should focus on dominant trees to maintain crown ratio and ensure wood quality.
The inflection point is an important feature of sigmoidal height-diameter (H-D) models. It is often cited as one of the properties favoring sigmoidal model forms. However, there are very few studies ...analyzing the inflection points of H-D models. The goals of this study were to theoretically and empirically examine the behaviors of inflection points of six common H-D models with a regional dataset. The six models were the Wykoff (WYK), Schumacher (SCH), Curtis (CUR), Hossfeld IV (HOS), von Bertalanffy-Richards (VBR), and Gompertz (GPZ) models. The models were first fitted in their base forms with tree species as random effects and were then expanded to include functional traits and spatial distribution. The distributions of the estimated inflection points were similar between the two-parameter models WYK, SCH, and CUR, but were different between the three-parameter models HOS, VBR, and GPZ. GPZ produced some of the largest inflection points. HOS and VBR produced concave H-D curves without inflection points for 12.7% and 39.7% of the tree species. Evergreen species or decreasing shade tolerance resulted in larger inflection points. The trends in the estimated inflection points of HOS and VBR were entirely opposite across the landscape. Furthermore, HOS could produce concave H-D curves for portions of the landscape. Based on the studied behaviors, the choice between two-parameter models may not matter. We recommend comparing several three-parameter model forms for consistency in estimated inflection points before deciding on one. Believing sigmoidal models to have inflection points does not necessarily mean that they will produce fitted curves with one. Our study highlights the need to integrate analysis of inflection points into modeling H-D relationships.
•Inflection points are important features of height-diameter models, but there is very little work examining them.•Two-parameter sigmoidal height-diameter models produced highly similar inflection points.•Three-parameter sigmoidal height-diameter models produced drastically different inflection points.•Some three-parameter sigmoidal height-diameter models produced concave curves without inflection points for certain tree species and parts of landscape.•We recommend integrating calculation of inflection points into model selection, analysis, and interpretation.
•Boosted Regression Trees (BRT) is embedded in Seemingly Unrelated Regression model.•Three-parameter Weibull function best fit diameter distribution of thinned stand.•Residual stand structures are ...the most influential predictors selected by BRT.•BRT is more robust than stepwise methods and leaps and bounds algorithm.•Partially linear model has the potentials for modeling diameter distribution.
Modeling stand diameter distribution is useful for many reasons such as predicting future product range. Parameter Prediction Method (PPM) is a commonly used method and has the advantage of interpreting parameters in terms of silvicultural practices and stand dynamics. Past PPM studies have seldom applied variable selection and cross-validation. The goal of this study was to embed a machine learning technique, Boosted Regression Trees (BRT), into PPM to address the knowledge gap. The PPM-BRT framework first applies BRT models to cross-validate and select stand attributes that are influential to the parameters of a probability density function (PDF) and then uses Seemingly Unrelated Regression (SUR) linear models to quantify the relationships. The framework was tested on Taiwania cryptomerioides thinning experiments. The three-parameter Weibull PDF had the best overall fit to the diameter distributions after thinning. The fitted BRT models explained about 76.9–86.8% of variations in the shape, scale, and location parameters. The final set of predictors selected by BRT that highly influenced the three parameters included number of years since thinning and the three moments of residual diameter distribution immediately after thinning. The SUR models showed that the shape and scale parameters were negatively associated with skewness of residual diameter distribution, but the location parameter was positively associated with it. Also, the three parameters were positively associated with number of years since thinning. This suggests that an intensive thinning from below results in a post-harvest diameter distribution that is more positively skewed, less variation in residual diameters, and larger minimum diameter. The diameter distribution would be less skewed and more heterogenous over time likely due to stem exclusion. Our study shows that BRT is more robust than stepwise regression. Future work could explore partially linear model for better integration of machine learning and parametric models.
Stand basal area (BA) is an important parameter, yet significantly influenced by initial planting density, in describing the developments of plantation forests. The effects of initial density on the ...growth of sugi (Cryptomeria japonica D. Don), an important plantation tree species in Taiwan, have seldom been studied. The goals of this study are to (1) model sugi stand BA growth under different initial densities, and (2) quantify model forecast accuracy. The data were from a spacing trial established in 1950 at the National Taiwan University Experimental Forest. The initial densities were 2500, 1111, 625, and 400 trees ha
−1
. Ten measurements were carried out up to stand age 65. Using a mixed-effects modeling approach, we fitted a Gompertz model with the initial density as the covariate to describe stand BA growth trends. Results showed that stand BA was the largest for stands with the highest initial density. The maximum current annual increment of the highest initial density was also the largest and reached earlier in stand development than that of the other spacings. Data up to stand age 60 were needed to yield a forecast accuracy within 6% for BA at stand age 65. This study showed that the fitted Gompertz model could adequately capture the general growth trends in stand BA despite unequal measurement intervals. The results agreed with other spacing trials in other regions and tree species, and the model accuracy was acceptable given the availability of data. The prediction accuracy results underlined the importance of long-term growth monitoring.
Measurement of tree attributes is important to collect information for forest management. Close-range photogrammetry with spherical panoramas has seen very little development and applications ...compared with aerial photography. This study develops methods to extract azimuth, horizontal distance, diameter at breast height, and upper stem diameters of individual trees from spherical panoramas based on (i) the trigonometry principle (TRIGO), (ii) the TRIGO corrected for terrain slope (TRIGOSLP), and (iii) the pinhole camera model (PINHOLE). Twenty-three horizontal point sample plots were randomly established in plantations in Taiwan, for a sample size of 486 trees. Results showed that tree azimuth was accurately and precisely estimated. TRIGO performed the worst in accuracy and precision for all other tree attributes. TRIGOSLP improved the results of TRIGO but had large estimation errors. PINHOLE achieved the best overall precision for all other tree attributes but was slightly inaccurate for estimating upper stem diameters. PINHOLE requires approaching a tree to attach a target of known size but has the ability to extract an almost continuous set of upper stem diameters from the tree, which could improve estimation of tree volume. Thus, PINHOLE could potentially be an alternative measurement system for hard-to-measure tree attributes.
Aim of study: Cluster plot designs are widely used in national forest inventory systems to assess current forest resources. By spreading subplots apart, a cluster plot could potentially capture a ...large variety of local plant species. This aspect has rarely been examined in the past. This study is conducted to understand how design factors of a cluster plot affect estimates of local plant species composition.Area of study: Two large census forest plots in Taiwan and Peninsular Malaysia over 25 ha with different species richness were used.Material and methods: Design factors of a cluster plot were plot configuration (PCONFIG), plot area (PAREA), cluster layout (CLAYOUT), and extent of ground area covered by a cluster (CEXTENT). Jaccard and Sørensen similarity indices were used to compare species compositional similarity between two cluster plot designs. A simulation study was carried out.Main results: Results were consistent among the study sites and similarity indices. PAREA, CLAYOUT, and CEXTENT notably influenced how species composition was sampled. Larger PAREA increased similarity in species composition between two cluster plot designs. Square and rectangle CLAYOUT had the most dissimilar species composition between them. Larger CEXTENT decreased similarity in species composition.Research highlights: We recommend that for CEXTENT ≤ 1000 m2 and PAREA ≤ 500 m2, a cluster plot of rectangle CLAYOUT is preferred for information gain. The study could potentially benefit forest managers designing cluster plots for plant diversity assessment.Keywords: Biodiversity assessment; composition similarity; national forest inventory; species diversity; sampling design; sampling efficiency.Abbreviation used: extent of ground area covered by a cluster (CEXTENT); cluster layout (CLAYOUT); Jaccard similarity index (JAC); plot area (PAREA); plot configuration (PCONFIG); Sørensen similarity index (SOR).
Desertification is a pressing issue in the dry Tarim River basin, which is under anthropogenic stresses. In this study, double sampling for stratification (DSS) is employed to inventory Populus ...euphratica Oliv. forests in the lower reaches of the Tarim River Basin in Xinjiang, China. The two objectives were evaluating DSS as a sampling technique for monitoring desertification and generating baseline information for permanent observation. Here, DSS consists of two phases: in phase 1, crown cover is observed on a large sample of plots on a high resolution satellite image, and these photo-plots are stratified into five crown cover strata. Phase 2 is a stratified random sample from these photo-plots and the sampled plots are field observed. Approximately 32% of the study area is without P. euphratica trees. As expected, estimated mean poplar tree density and basal area increase with crown cover. DSS takes advantages of stratification (fieldwork efficiency and statistical precision) without the need for a priori strata delineation. It proves feasible for inventory the sparse poplar population and holds promise for the assessment of trees outside the forest, where density varies considerably and pre-stratification is intractable. It can be integrated into permanent observation systems for monitoring vegetation changes.
•We model habitat of rare rodent species with Generalized Linear Mixed Models.•We compare results from GLMMs to models without mixed effects.•GLMMs predicted stronger and sometimes contrary habitat ...effects compared to GLMs.•Overdispersed Poisson GLMM provided the most consistent fit to rare species.•Rare species need not be excluded from microhabitat association modeling.
Knowledge about the relationship between habitat structure and abundance of a target species facilitates biodiversity conservation in managed forests. However, modeling the relationship for infrequent small mammal species in silvicultural experiments introduces the challenge of excessive zero counts and complex hierarchical sampling. A common solution has been to ignore infrequent species. The goal of this study was to model microhabitat associations of infrequently captured forest floor small mammal species with Bayesian models that accounted for subsampling and the blocking design of a large-scale variable-retention harvest experiment. Poisson, negative binomial and overdispersed Poisson Generalized Linear Mixed Models (GLMMs) were fitted to data for three small mammal species with different rates of capture. Shrew–mole (Neurotrichus gibbsii) and Keen’s deer mouse (Peromyscus keeni) were the two infrequent species and southern red-backed vole (Myodes gapperi) was the frequent species selected for modeling. Capture rate was predicted from variables representing vegetation structure, and results were compared to corresponding Generalized Linear Models (GLMs). GLMMs predicted stronger and sometimes contrary effects of vegetation structure with wider confidence intervals compared to GLMs. The overdispersed Poisson GLMM provided the most consistent and adequate fit to infrequent species. Capture rate of the shrew–mole was found to be negatively associated with tall shrub cover and coarse woody debris volume. Similarly, capture rate of Keen’s deer mouse was negatively associated with herb cover and coarse woody debris volume. Finally, captures of southern red-backed vole was associated negatively with herb cover and coarse woody debris volume but positively associated with vertical complexity of overstory vegetation. With correct GLMM specification, statistical inferences of habitat predictors were more reliable as autocorrelation between samples was properly accounted for and valid standard errors were estimated. Furthermore, the GLMMs in this study fitted capture rates of infrequent species well and produced admissible results on the association of these species to microhabitat features. Infrequent species need not be excluded from analysis; in fact, inclusion of these species is crucial to conservation of species diversity by designing silvicultural treatments that produce or protect suitable habitat.
Sustainable forest management needs information on spatial distribution of species richness. The objectives of this study were to understand whether knowledge, method, and effort of a rapid ...assessment affected accuracy and consistency in mapping species richness. A simulation study was carried out with nine 25-50 ha census plots located in tropical, subtropical, and temperate zones. Each forest site was first tessellated into non-overlapping cells. Rapid assessment was conducted in all cells to generate a complete coverage of proxies of the underlying species richness. Cells were subsampled for census, where all plant individuals were identified to species in these census cells. An artificial neural network model was built using the census cells that contain rapid assessment and census information. The model then predicted species richness of cells that were not censused. Results showed that knowledge level did not improve the accuracy and consistency in mapping species richness. Rapid assessment effort and method significantly affected the accuracy and consistency. Increasing rapid assessment effort from 10 to 40 plant individuals could improve the accuracy and consistency up to 2.2% and 2.8%, respectively. Transect reduced accuracy and consistency by up to 0.5% and 0.8%, respectively. This study suggests that knowing at least half of the species in a forest is sufficient for a rapid assessment. At least 20 plant individuals per cell is recommended for rapid assessment. Lastly, a rapid assessment could be carried out by local communities that are familiar with their forests; thus, further supporting sustainable forest management.