Inventory Plots Research Articles

High regeneration may not contribute to the forest's carbon storage: a case study in the mangrove forest of Rajang-Belawai-Paloh delta, Sarawak.

The alarming rate of the mangrove ecosystem loss poses a threat of losing valuable carbon sinks. This study was conducted to (i) determine the growth structure in different vegetation types and (ii) compare the aboveground biomass (AGB) and carbon storage in different vegetation types. The study was conducted at four vegetation types within the Rajang-Belawai-Paloh delta i.e., Matured Bakau-Berus Forest (MBBF), Bakau-Nipah Forest (BNF), Regenerating Forests (Debris pile) [RF-D], and Regenerating Forests (Machinery track) [RF-M]. Inventory plots (20 m × 20 m) are systematically located along the main waterways and smaller rivers/streams. Trees (≥ 5 cm diameter-at-breast height [DBH]), seedlings (< 2-cm stem diameter), and saplings (2-4.9-cm stem diameter) were measured. The trend of total trees per hectare is found to be decreasing across the least disturbed vegetation (MBBF) to the most disturbed vegetation (RF-M). The trends of total seedlings and saplings per hectare are found to be going upwards from the least disturbed vegetation to the most disturbed vegetation. Kruskal-Wallis H-test showed that there is a significant difference in the AGB and carbon storage between different vegetation types, χ2(2) = 43.98, p = 0.00 with the highest mean rank AGB and carbon storage in BNF (612.20 t/ha) and lowest in RF-M (287.85 t/ha). It can be concluded that although the most disturbed vegetations have higher regeneration, it may not contribute to the forest's carbon storage The naturally regenerated seedlings may not grow beyond the sapling stage unless sustainable forest management is conducted to ensure survivability and growth.

Computational tools for assessing forest recovery with GEDI shots and forest change maps

Tropical secondary forests are ecosystems of critical importance for protecting biodiversity, buffering primary forest loss, and sequestering atmospheric carbon. Monitoring growth and carbon sequestration in secondary forests is difficult, with inventory plots sampling <0.001% of secondary forests. The Global Ecosystem Dynamics Investigation (GEDI), a space-borne LiDAR sampler, provides billions of aboveground carbon density (ACD) estimates across the tropics. We fuse these carbon density estimates with a time series of forest change maps to identify their age since last deforestation and thus estimate the average rate of carbon sequestration in secondary forests across the Amazon biome. To our knowledge, this is the first estimate of these rates made using the new GEDI dataset. Moreover, this paper addresses key statistical and computational challenges of GEDI data fusion and analysis. We propagate both GEDI ACD and geolocation uncertainty to the regrowth rate estimate through a Monte Carlo approach, and we handle heteroskedasticity, outliers, and spatial autocorrelation using robust statistical methods. The large size of the GEDI dataset combined with the proposed Monte Carlo bootstrap can be highly computationally intensive, with a naive implementation taking over a month to complete. Nevertheless, we demonstrate the feasibility of our method by developing optimized open-source code that performs this computation on the 151 million quality-filtered GEDI shots available for the Amazon biome from April 2019–August 2021 in under 25 min in benchmark tests. By resolving these statistical and computational challenges with an efficient open-source pipeline, we create a standard approach that can be used more broadly in any work seeking to combine the GEDI dataset with high-resolution classification maps. Using this approach, we identify approximately 23, 000 GEDI samples of regrowing forest at least 60 m × 60 m wide across the Amazon biome and estimate a carbon sequestration rate of 1.86 MgC/ha/yr with a 95% empirical confidence interval of 1.75–1.97 MgC/ha/yr, with rates varying from 1.27 to 1.99 MgC/ha/yr across smaller subregions.

FORMS: Forest Multiple Source height, wood volume, and biomass maps in France at 10 to 30 m resolution based on Sentinel-1, Sentinel-2, and Global Ecosystem Dynamics Investigation (GEDI) data with a deep learning approach

Abstract. The contribution of forests to carbon storage and biodiversity conservation highlights the need for accurate forest height and biomass mapping and monitoring. In France, forests are managed mainly by private owners and divided into small stands, requiring 10 to 50 m spatial resolution data to be correctly separated. Further, 35 % of the French forest territory is covered by mountains and Mediterranean forests which are managed very extensively. In this work, we used a deep-learning model based on multi-stream remote-sensing measurements (NASA's Global Ecosystem Dynamics Investigation (GEDI) lidar mission and ESA's Copernicus Sentinel-1 and Sentinel-2 satellites) to create a 10 m resolution canopy height map of France for 2020 (FORMS-H). In a second step, with allometric equations fitted to the French National Forest Inventory (NFI) plot data, we created a 30 m resolution above-ground biomass density (AGBD) map (Mg ha−1) of France (FORMS-B). Extensive validation was conducted. First, independent datasets from airborne laser scanning (ALS) and NFI data from thousands of plots reveal a mean absolute error (MAE) of 2.94 m for FORMS-H, which outperforms existing canopy height models. Second, FORMS-B was validated using two independent forest inventory datasets from the Renecofor permanent forest plot network and from the GLORIE forest inventory with MAE of 59.6 and 19.6 Mg ha−1, respectively, providing greater performance than other AGBD products sampled over France. Finally, we compared FORMS-V (for volume) with wood volume estimations at the ecological region scale and obtained an R2 of 0.63 with an MAE of 30 m3 ha−1. These results highlight the importance of coupling remote-sensing technologies with recent advances in computer science to bring material insights to climate-efficient forest management policies. Additionally, our approach is based on open-access data having global coverage and a high spatial and temporal resolution, making the maps reproducible and easily scalable. FORMS products can be accessed from https://doi.org/10.5281/zenodo.7840108 (Schwartz et al., 2023).

Above ground tree biomass modeling using machine learning algorithms in western Terai Sal Forest of Nepal

The monitoring of forest biomass is a crucial biophysical parameter in forest ecosystems, as it provides valuable information for managing forests sustainably and tracking carbon circulation statistics. To achieve sustainable forest management, it is essential to monitor and study forest resources, particularly biomass. This study aimed to model above ground tree biomass (AGTB) using Machine Learning Algorithms (MLAs) in the western terai Sal forest of Nepal. AGTB was calculated using a systematic inventory sample plot, while spectral and textural variables were processed and masked for the study area using Sentinel-2A satellite imagery. Three MLAs namely support vector machine (SVM), random forest (RF), and stochastic gradient boosting (SGB), were employed for modeling with eight categorized variable datasets. Among the MLAs, the RF algorithm with a combination of gray-level co-occurrence matrix (GLCM) and raw bands (RB) dataset variable demonstrated the best performance, with a low RMSE value of 78.81 t ha−1 in the test data. However, the AGTB range from this model ranged from 118.34 to 425.97 t ha−1. The study found that traditional indices, raw bands, and GLCM texture from near-infrared were important variables for AGTB. Nevertheless, the RF algorithm and the dataset combination of GLCM plus raw bands (RB) exhibited excellent performance in all model runs. Thus, this pioneering study on comparative MLAs-based AGTB assessment with multiple datasets variables can provide valuable insights for new researchers and the development of novel approaches for biomass/carbon estimation techniques in Nepal.

Climate Change Impacts on Species Composition and Floristic Regions in Thailand

Tropical forests are vulnerable to climate change including increased temperatures and changes to rainfall variation. Here, we use Thailand as a case study for assessing the impacts of the shared socio-economic pathway and climate scenarios on changes to the distribution and extent of floristic regions. To address this question, we assigned floristic regions based on a structured cluster analysis of modeled species ranges, and evaluated how those regions respond under scenarios of climate change. A total of 201 plant species with sufficient occurrence data obtained from the systematic forest inventory plots across the country and global datasets were chosen for distribution modeling. Environmental variables, including soils, topography and bioclimatic variables were compiled to model both current and 2050 distributions. Potential floristic regions were classified using a clustering algorithm on the pixel-wise species compositions—resulting in 12 floristic regions representative of both current climate species compositions and projected future species assemblages. Five floristic regions are projected to experience little change in their geographic distribution, while the remainder are projected to be substantially displaced spatially. Additionally, two of the identified floristic regions are not well represented in protected areas—with less than 50% of the current geographic distribution in each region in some form of protected status.

Estimated mangrove carbon stocks and fluxes to inform MRV for REDD+ using a process-based model

Mangrove forests are important due to their strong capability for carbon storage, especially in soils. Understanding carbon dynamics in these forests is fundamental to estimate their roles in carbon storage and mitigating climate change. This study used a process-based model, MCAT-DNDC, to assess mangrove carbon sequestration and fluxes at a 30-m spatial resolution in three African countries, Gabon, Mozambique and Tanzania. The simulated above- and below-ground biomass at inventory plots in each country was approximate to actual observations with mean errors <5% for aboveground biomass and <8% for belowground biomass, indicating that the MCAT-DNDC model can be a useful tool for assessing mangrove carbon storage and fluxes. The results from assessing mangrove carbon storage and fluxes for the three countries showed that the mangroves in these countries are large carbon pools, they export large amounts of dissolved and particulate carbon components to riverine and oceanic ecosystems, and they bury a large amount of carbon in soils. However, soil-borne greenhouse gases CO2, CH4 and N2O fluxes from mangrove forest lands were low. There were large differences in all mangrove carbon components among the mangroves in the three countries, indicating that regional or global mangrove carbon stocks estimated using a process-based model may be better than the extrapolations using limited inventories.

Effects of site conditions on regeneration of African mahogany (Khaya anthotheca) in a semi‐deciduous East African tropical rainforest

Restoration of high‐value timber trees in logged tropical forests is indispensable as a conservation strategy and for the enhancement of ecosystem services. Khaya anthotheca is a mahogany tree species of major restoration importance in Uganda. However, the factors affecting the natural regeneration success of this species are poorly understood. We determined the effect of selected site conditions on the abundance of K. anthotheca regeneration in the Budongo Forest Reserve. Two hundred and four 1,000 m2 inventory plots, each nested with four 4 m2 subplots, were systematically established in logged and unlogged sites within the study area. All trees greater than or equal to 10 cm DBH and natural regeneration (&lt;10 cm DBH) were identified and recorded in the plots and subplots, respectively. We characterized plots by the density of conspecifics, light availability, ground cover of undergrowth, litter depth, soil pH, and basal area. Regression models were used to evaluate the variables that influenced regeneration abundance and occurrence. The density of adult conspecifics and soil pH were important. Regeneration abundance and occurrence increased with increasing density of adult conspecifics and decreased with increasing soil pH. Our results highlight that in addition to seed availability, soil pH was important for successful regeneration. This is one of the first K. anthotheca‐specific studies to evaluate more than a few environmental factors influencing the density of its naturally established seedlings in tropical African forests. The abundance and distribution of adult conspecifics in the canopy and soil pH should be considered in K. anthotheca restoration plans.

It’s about Time: A Method for Estimating Wildfire Arrival and Weather Conditions at Field-Sampled Locations

Weather conditions at the time of wildfire front arrival strongly influence fire behavior and effects, yet few methods exist for estimating weather conditions more spatio-temporally resolved than coarse-grain (e.g., 4 km) daily averages. When a fire front advances rapidly and weather conditions are heterogeneous over space and time, greater spatio-temporal precision is required to accurately link fire weather to observed fire effects. To identify the influence of fire weather on fire effects observed across a sample of existing forest inventory plots during a wind-driven megafire event in the US Pacific Northwest, we explored and compared three methods for estimating time of fire arrival and the wind speed at that arrival time for each plot location. Two methods were based on widely used, remotely sensed active fire data products with dissimilar spatial and temporal resolutions. The third and preferred method, Modeled-Weather Interpolated Perimeters (MoWIP), is a new approach that leveraged fine-grained (1.3 km, hourly) wind speed and direction from modeled fire weather to guide interpolation of aerial infrared-detected (IR) operational perimeters, subdividing the time intervals defined by sequential IR perimeters into quartile intervals to enhance temporal resolution of predicted fire arrival times. Our description of these fire arrival “time stamp” methods and discussion of their utility and shortcomings should prove useful to fire scientists, ecologists, land managers, and future analyses seeking to link estimated fire weather and observed fire effects.

Mapping site index in coniferous forests using bi-temporal airborne laser scanning data and field data from the Swedish national forest inventory

Recent advancements in remote sensing of forests have demonstrated the capabilities of three-dimensional data acquired by airborne laser scanning (ALS) and, consequently, have become an integral part of enhanced forest inventories in Northern Europe. In Sweden, the first national laser scanning revolutionised forest management planning through low-cost production of large-scale and spatially explicit maps of forest attributes such as basal area, volume, and biomass, compared to the earlier practice based on field survey data. A second scanning at the national level was launched in 2019, and it provides conditions for the estimation of height growth and site index. Accurate and up-to-date information about site productivity is relevant for planning silvicultural treatments and for the prognosis of forest status and development over time. In this study, we explored the potential of bi-temporal ALS data and other auxiliary information to predict and map site productivity by site index according to site properties (SIS) of Norway spruce (Picea abies (L.) Karst) and Scots pine (Pinus sylvestris L.) in even-aged stands in Sweden. We linked ground survey data of SIS from more than 11,500 plots of the Swedish National Forest Inventory (NFI) to bi-temporal ALS data to predict and map site index using an area-based method and two regression modelling strategies: (1) a multiple linear regression (MLR) model with an ordinary least-squares parameter estimation method, and (2) a non-parametric random forests (RF) model optimised for hyper-parameter tuning. For model development, permanent plots were used, whereas the validation was done on the temporary plots of the Swedish NFI and an independent stand-level dataset. Species-specific models were developed, and the root mean square error (RMSE) metric was used to quantify the residual variability around model predictions. For both species, the MLR model gave precise and accurate estimates of SIS. The RMSE for SIS predictions was in the range of 1.96 – 2.11 m, and the relative RMSE was less than 10 % (7.68 – 9.49 %) of the reference mean value. Final predictors of site index include metrics of 90th percentile height and annual increment in the 95th percentile height, altitude, distance to coast, and soil moisture. Country-wide maps of SIS and the corresponding pixel-level prediction errors at a spatial resolution of 12.5 m grid cells were produced for the two species. Independent validations show the site index maps are suitable for use in operational forest management planning in Sweden.

Relationships among livestock, structure, and regeneration in Chilean Austral Macrozone temperate forests

A Macrozone is a socioecological region with shared geographic and demographic characteristics. Within the Chilean Austral Macrozone (43° to 56° SL), the native temperate forests serve as a crucial resource, offering multiple ecosystem services to local communities. These forests significantly support animal husbandry practices involving cattle, horses and sheep. However, introducing these exotic species affects natural regeneration and compromises their long-term sustainability. This study proposes a new classification of the temperate forests in the Chilean Austral Macrozone based on structure parameters and determine their relationships with animal husbandry and natural regeneration. Data were obtained from Chile's National Forest Inventory (NFI) (2001–2010), including 195 inventory plots (500 m2) with 21 tree species. We redefined the forest categories described in NFI according to the proportional basal area of each tree species at each plot. We used two levels of analysis: forest composition (Level 1), which includes general categories such as mono-specific (dominated by a single tree species), bi-specific (dominated by two tree species), and multi-specific forests (dominated by multiple tree species), and forest type (Level 2), which includes specific species (e.g., Nothofagus pumilio) or species groups (e.g., N. pumilio - N. dombeyi) with economic relevance. We evaluated the data using univariate and multivariate analyses. We found 18 forest types in the Austral Macrozone, in contrast to the three traditionally recognized forest types used in the NFI (e.g., lenga, evergreen, coihue de Magallanes). Livestock was observed in all forest types, where Nothofagus forests showed that regeneration in N. antarctica and N. pumilio were higher with livestock than without livestock breeding (Hedges' g > 0.51). The natural regeneration of the studied forests was influenced by animal husbandry, environmental variables (bioclimatic and topographic factors), and forest structure. Our data suggested the importance of using more forest types than the three classics to generate tools or recommendations that are more focused on the particularities of each one. The classification must be based on forest parameters obtained during NFI. The proposed forest type classification reflects the complexity and richness of the forests in a better way, which could improve forest management and the development of public policies related to climate change and sustainability. Finally, although livestock pressure was observed in all forest types, the impact over some areas (e.g., N. antarctica) needs special attention in the forest management and conservation planning for the Chilean Austral Macrozone.

Relationships between juvenile tree survival and tree density, shrub cover and temperature vary by size class based on ratios of abundance

Global change drivers are altering forest dynamics, yet how these factors influence tree survival across early developmental stages (i.e., seedling to recruited sapling) over large geographies is not well understood. We developed a novel approach to evaluate controls on seedling and sapling survival. This approach was demonstrated on a set of systematic forest inventory plots across the northeastern USA in which seedlings were tallied within six height classes, allowing for a detailed assessment of the stages at which demographic bottlenecks in juvenile tree development are often observed. Forest inventory subplots containing a study species were divided into overlapping bins along an environmental or ecological gradient, and ratios of abundance between successive size classes were used to infer relative survival rates. Relationships between 10 common tree species and tree density, shrub cover, and mean annual temperature were assessed. As seedling height class increased, we observed shifts from positive to negative associations with shrub cover and large tree density. Our results suggest that observed patterns of sapling and tree abundance may belie complex and sometimes opposing influences on seedling survival that are important to quantify when predicting and managing for successful tree recruitment and future canopy tree composition.

Comparing harmonic regression and GLAD Phenology metrics for estimation of forest community types and aboveground live biomass within forest inventory and analysis plots

The Landsat multispectral time series is a valuable source of moderate spatial resolution data to support forest mapping and monitoring tasks. Using United States Department of Agriculture (USDA) Forest Service Forest Inventory and Analysis (FIA) plots within the states of Michigan, Oregon, and West Virginia, two methods to summarize time series observations, harmonic regression coefficients and Global Land Analysis & Discovery (GLAD) Phenology metrics, are compared for predicting forest community type, total aboveground live biomass (AGLBM), and species-specific AGLBM. Harmonic regression coefficients, which provided mean overall accuracies (OAs) between 62.8% and 73.1% and map image classification efficacies (MICEs) varying between 0.455 and 0.566 for the three studied states and calculated using multiple model replicates, generally provided better predictive performance for differentiating forest community types in comparison to GLAD phenology metrics. However, differences were not always statistically significant. DTM-derived terrain variables improved classification performance in some landscapes when using machine learning, random forest classification (for example, the highest obtained mean OA of 78.5% (MICE = 0.557) was obtained for West Virginia when using the combined harmonic regression and terrain variables). For the regression-based estimation of total and species-specific AGLBM, the spectral data used and the incorporation of terrain variables had less of an impact on model performance. Further, the regression models in Oregon provided larger mean R-squared values (0.452 to 0.490) in comparison to those in Michigan and West Virginia, where all R-squared values were lower than 0.200, suggesting that AGLBM prediction may be more or less challenging in different landscapes depending on forest characteristics, terrain, management practices, and disturbance histories.

Modeling Height–Diameter Relationship Using Artificial Neural Networks for Durango Pine (Pinus durangensis Martínez) Species in Mexico

The total tree height (h) and diameter at breast height (dbh) relationship is an essential tool in forest management and planning. Nonlinear mixed effect modeling (NLMEM) has been extensively used, and lately the artificial neural network (ANN) and the resilient backpropagation artificial neural network (RBPANN) approach has been a trending topic for modeling this relationship. The objective of this study was to evaluate and contrast the NLMEN and RBPANN approaches for modeling the h-dbh relationship for the Durango pine species (Pinus durangensis Martínez) for both training and testing datasets in a mixed-species forest in Mexico. The knowledge of this relationship is important for forest management and planning in Mexican Forestry. The total dataset considered 1000 plots (each plot 0.10 ha) (11,472 measured trees) randomly selected from 14,390 temporary forest inventory plots and the dataset was randomly divided into two parts: 50% for training and 50% for testing. An unsupervised clustering analysis was used to group the dataset into 10 cluster-groups based on the k-means clustering method. An RBPANN was performed for tangent hyperbolicus (RBPANN-tanh), softplus (RBPANN-softplus), and logistic (RBPANN-logistic) activation functions in the cross product of the covariate or neurons and the weights for the ANN analysis. Also, a different vector of hidden layers was used for training of ANNs. For both training and testing, 10 classical statistics (e.g., RMSE, AIC, BIC, and logLik) were computed for the residual values and to assess the approaches for the h-dbh relationship. For training and testing, the ANN approach outperformed the NLMEM approach, and the RBPANN-tanh had the best performance in both the training and testing of ANNs.

Density-dependent mortality models for mono- and multi-species uneven-aged stands: The role of species mixture

The self-thinning law describes the progressive mortality of growing trees in crowded even-aged stands because of competition. In uneven-aged stands, however, density-dependent mortality is relatively poorly known. We developed density-dependent self-thinning models for mono-species and multi-species uneven-aged stands. Three models for mono-species Picea abies, Abies alba and Fagus sylvatica stands, four models for mixed conifer/broadleaf stands and a model for single tree selection stands as a specific type of uneven-aged forest were parametrized by Quantile Regression (QR) and Stochastic Frontier Analysis (SFA) using data from n = 31,612 forest inventory plots in Slovenia. The average slope of the self-thinning line in uneven-aged stands of −1.422 was statistically significantly lower than Reineke’s slope. Maximum stand density was higher in mono-species stands than in multi-species stands. The higher the proportion of beech in a stand, the higher the intercept and slope of the maximum stand density line. In single tree selection stands, the intercept and the slope of the maximum stand density line were highest and statistically different from all other types of uneven-aged forests. Species mixture had an impact on the slope and intercept of the self-thinning line. The slopes of the self-thinning lines differed between the studied types by up to 36% and 69% in the SFA and QR models, respectively, while the intercepts differed by up to 13% and 24%, respectively. However, maximum stand densities differed significantly only between compositionally very dissimilar stand types, such as Fagus sylvatica- or Picea abies-dominated mixtures versus other broadleaf mixtures (e.g., Quercus petraea, Carpinus betulus). The SFA predictions of maximum stand densities were systematically lower and less varying than the QR predictions, but not statistically different. The models can be used to constrain individual tree growth models in uneven-aged forests and to estimate the onset of inter-tree competition in irregular stands.

Improving spatial predictions of Eucalypt plantation growth by combining interpretable machine learning with the 3-PG model

Accurate predictions of forest plantation growth provide forest managers with improved forest inventory estimates, forest valuation, and timely harvest schedules. Forest process-based models are increasingly used for quantifying current and potential productivity, yield gaps, and factors limiting growth, such as climate variability, soil characteristics, and water deficit. Improvements in the availability and resolution of spatial and temporal data combined with advancements in machine learning algorithms provide new opportunities to improve model predictions. This study shows how interpretable machine learning (ML) can be used to independently predict site soil fertility rating (FR) and incorporate these results into the 3-PG forest process-based model to accurately predict plantation growth. Four ensemble decision tree machine learning models—random forest trees, extremely randomized trees, gradient boost, and XG boost—were trained and compared using spatial cross-validation across the study area. FR predictions were estimated in relation to the influencing soil type and terrain characteristics, and interpretable ML methods were used to show how input feature permutations may relate to the soil fertility predictions. The results show the explanatory variables are similar across the selected ML models, with the strongest influencing variables being water leaching index, site aspect, and the silt and sand soil texture properties. The extremely randomized tree models showed the overall best performance, with only a small variation in performance across the four ML models. The method was applied to Eucalyptus nitens plantations covering over 63,000 ha in north-west Tasmania, Australia. The results using the predicted FR spatial grid with 3-PG show a strong correlation with observed growth for tree diameter and stand volume (R2 tree diameter at breast height = 0.97, RMSE = 0.85 m; R2 stand volume = 0.96, RMSE = 23.1 m3 ha−1) obtained from 161 permanent sample inventory plots ranging from 3 to 31 years old. This method has practical utility for other study sites to calibrate forest plantation soil fertility rating, in both the spatial and point-scale 3-PG model, where spatial data of soil characteristics are available. The derived soil fertility grid can provide valuable insights into the spatial variability of soil fertility in unknown areas.

MADERA: A standardized Pan-Amazonian dataset for tropical timber species.

We compiled and presented a dataset for all timber species reported in the Amazon region from all nine South American Amazonian countries. This was based on official information from every country, as well as from two substantial scientific references. We verified the standard taxonomic names from each individual source, using the Taxonomic Name Resolution Service (TNRS) and considered all Amazonian tree species with diameter at breast height (DBH) ≥10 cm. We also obtained estimates of the current population size for most species from a published approach based on data from 1900 tree inventory plots (1-ha each) distributed across the Amazon region and part from the Amazon Tree Diversity Network (ATDN). We then identified the hyperdominant timber species. In addition, we overlapped our timber species list with data for species that are used for commercial purposes, according to the International Tropical Timber Organization (ITTO), the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) and the International Union for Conservation of Nature (IUCN) taxa assessment and Red List categories. Finally, we also included IUCN Red List categories based on combined deforestation, and climate change scenarios for these species. Our final Amazonian timber species dataset contains 1112 unique species records, which belong to 337 genera and 72 families from the lowland Amazonian rainforest, with associated information related to population, conservation, and trade status of each species. The authors of this research expect that the information provided will be useful to strengthen the public forestry policies of the Amazon countries, inform ecological studies, as well for forest management purposes. The data are released under the Creative Commons Attribution 4.0 International license.

Decision-Tree Application to Predict and Spatialize the Wood Productivity Probabilities of Eucalyptus Plantations

Brazil is one of the world’s wood short-fiber producers, cultivating 7.5 million hectares of eucalypt trees. Foresters and resource managers often face difficulties in surveying reliable Eucalyptus productivity levels for the purpose of purchasing and prospecting lands. Spatial data science (DS) and machine learning (ML) provide powerful approaches to make the best use of the large datasets available today. Agriculture has made great use of these approaches, and in this paper, we explore how forestry can benefit as well. We hypothesized that both DS and ML techniques can be used to improve Eucalyptus productivity zoning based on multiple operational datasets of tree growth and environment. Based on more than 12,000 permanent forest inventory plots of commercial Eucalyptus plantations and the climate, soil, and altitude variables associated with them, a supervised ML approach was adjusted to model the forest plantation productivity. A multi-tuning of the decision-tree (DT) algorithm hyperparameters was prepared to yield 450 DT models, with a better one delivering an RMSE of 53.5 m3 ha−1, split in 35 terminal nodes, here interpreted as Eucalyptus productivity zones. The DT model showed an optimum performance index of 0.83, a coefficient of determination of 0.91, a root mean squared error of 12.3 m3 ha−1, and a mean absolute percentage error only of 3.1% in predicting the testing dataset throughout the study area. The DT rule set was interpreted in a user-friendly table and was prepared to classify any location within the study area in each one of the 35 productivity zones based on the required environment variables of the DT algorithm. The high quality of the model obtained made it possible to spatialize the DT rules, providing a reliable cartographic visualization of the probability levels of true Eucalyptus productivity for a huge region of forest-based industries in Brazil. These data-science techniques also provided a yield gap analysis using a very down-to-earth approach. We estimated a yield gap by an amount of 4.2 × 107 m3, representing a few more than 113,000 ha, or 15% of the current forest base. This is the amount of avoided area expansion to accumulate the same wood stock in case the productivity is raised to the attainable level in each zone. This present study provided deeper analysis and reproducible tools to manage forest assets sustainably.

Plot-level estimates of aboveground biomass and soil organic carbon stocks from Nepal’s forest inventory

Given the contribution of deforestation and forest degradation to the global carbon cycle, forest resources are critical to mitigating the global climate change effects. Improved forest monitoring across different biomes is important to understand forest dynamics better and improve global projections of future atmospheric CO2 concentration. Better quantification of the forest carbon cycle advances scientific understanding and informs global negotiations about carbon emissions reduction. High-quality estimates of forest carbon stocks are currently scarce in many developing countries. Here, we present the most comprehensive georeferenced data set to date of plot-level forest carbon estimates for Nepal. Based on field observations from Nepal’s national forest inventory of 2010–2014; the data set includes estimates for two major forest carbon pools, aboveground biomass (AGB) and soil organic carbon (SOC) stocks from 2,009 and 1,156 inventory plots, respectively. The dataset fills an important knowledge gap about forest carbon stocks in the Central Himalayas, a region with highly heterogeneous environmental conditions and rich biodiversity that is poorly represented in existing global estimates of forest carbon.

Examining the interactive effects of neighborhood characteristics and environmental conditions on height-to-diameter ratio of Chinese fir based on random forest

In forest dynamics modeling and estimation, the nonlinear relationship between response and explanatory variables is often concerned, while the interaction among variables is neglected. Height-to-diameter ratio (HDR) is an important indicator for evaluating tree stability and performance. Based on the data from 112 forest inventory plots across Hunan province, China, we modeled the HDR of Chinese fir (Cunninghamia lanceolata (Lamb.) Hook) by using random forest and examined the interactive effects of neighborhood variables (neighborhood competition, density and mean diameter) and environmental conditions (dominant height, topography and climate) on the HDR of Chinese fir. Neighborhood competition contributed the most to HDR variation and affected HDR positively. Dominant height and elevation also had a strong influence on HDR. Furthermore, the changes in HDR were jointly dominated by neighborhood features and environmental conditions. For example, HDR varied from 50.5 to 80.3 under the interactive influence of neighborhood competition and dominant height, but varied from 56.8 to 70.2 under the influence of neighborhood competition alone. The relationship between HDR and neighborhood variables might be mediated by environmental conditions, while neighborhood variables did not change the overall trend of HDR responses to environmental factors. When the dense Chinese fir stand is located at low elevation (<500 m) or rich sites (>20 m), frequent and systematic thinning should be carried out in the early stages of forest development. Our findings can provide guidance for managing Chinese fir forests to improve their stability.

Declining tree growth resilience mediates subsequent forest mortality in the US Mountain West.

Climate change-triggered forest die-off is an increasing threat to global forests and carbon sequestration but remains extremely challenging to predict. Tree growth resilience metrics have been proposed as measurable proxies of tree susceptibility to mortality. However, it remains unclear whether tree growth resilience can improve predictions of stand-level mortality. Here, we use an extensive tree-ring dataset collected at ~3000 permanent forest inventory plots, spanning 13 dominant species across the US Mountain West, where forests have experienced strong drought and extensive die-off has been observed in the past two decades, to test the hypothesis that tree growth resilience to drought can explain and improve predictions of observed stand-level mortality. We found substantial increases in growth variability and temporal autocorrelation as well declining drought resistance and resilience for a number of species over the second half of the 20th century. Declining resilience and low tree growth were strongly associated with cross- and within-species patterns of mortality. Resilience metrics had similar explicative power compared to climate and stand structure, but the covariance structure among predictors implied that the effect of tree resilience on mortality could partially be explained by stand and climate variables. We conclude that tree growth resilience offers highly valuable insights on tree physiology by integrating the effect of stressors on forest mortality but may have only moderate potential to improve large-scale projections of forest die-off under climate change.