Fake legal logging in the Brazilian Amazon

How to discriminate wood of CITES-listed tree species from their look-alikes: using an attention mechanism with the ResNet model on an enhanced macroscopic image dataset

Introduction

Global illegal trade in timbers is a major cause of the loss of tree species diversity. The Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) has been developed to combat the illegal international timber trade. Its implementation relies on accurate wood identification techniques for field screening. However, meeting the demand for timber field screening at the species level using the traditional wood identification method depending on wood anatomy is complicated, time-consuming, and challenging for enforcement officials who did not major in wood science.

Methods

This study constructed a CITES-28 macroscopic image dataset, including 9,437 original images of 279 xylarium wood specimens from 14 CITES-listed commonly traded tree species and 14 look-alike species. We evaluated a suitable wood image preprocessing method and developed a highly effective computer vision classification model, SE-ResNet, on the enhanced image dataset. The model incorporated attention mechanism modules [squeeze-and-excitation networks (SENet)] into a convolutional neural network (ResNet) to identify 28 wood species.

Results

The results showed that the SE-ResNet model achieved a remarkable 99.65% accuracy. Additionally, image cropping and rotation were proven effective image preprocessing methods for data enhancement. This study also conducted real-world identification using images of new specimens from the timber market to test the model and achieved 82.3% accuracy.

Conclusion

This study presents a convolutional neural network model coupled with the SENet module to discriminate CITES-listed species with their look-alikes and investigates a standard guideline for enhancing wood transverse image data, providing a practical computer vision method tool to protect endangered tree species and highlighting its substantial potential for CITES implementation.

Sustainable Development versus Extractivist Deforestation in Tropical, Subtropical, and Boreal Forest Ecosystems: Repercussions and Controversies about the Mother Tree and the Mycorrhizal Network Hypothesis

This research reviews the phenomenon of extractive deforestation as a possible trigger for cascade reactions that could affect part of the forest ecosystem and its biodiversity (surface, aerial, and underground) in tropical, subtropical, and boreal forests. The controversy and disparities in criteria generated in the international scientific community around the hypothesis of a possible link between "mother trees" and mycorrhizal networks in coopetition for nutrients, nitrogen, and carbon are analyzed. The objective is to promote awareness to generate more scientific knowledge about the eventual impacts of forest extraction. Public policies are emphasized as crucial mediators for balanced sustainable development. Currently, the effects of extractive deforestation on forest ecosystems are poorly understood, which requires caution and forest protection. Continued research to increase our knowledge in molecular biology is advocated to understand the adaptation of biological organisms to the new conditions of the ecosystem both in the face of extractive deforestation and reforestation. The environmental impacts of extractive deforestation, such as the loss of biodiversity, soil degradation, altered water cycles, and the contribution of climate change, remain largely unknown. Long-term and high-quality research is essential to ensure forest sustainability and the preservation of biodiversity for future generations.

Wood volume is overestimated in the Brazilian Amazon: Why not use generic volume prediction methods in tropical forest management?

The Amazon has a range of species with high potential for sustainable timber harvesting, but for them to be utilized globally, the merchantable wood volume must be accurately quantified. However, since the 1950s, inadequate methods for estimating merchantable timber volumes have been employed in the Amazon, and Brazilian Government agencies still require some of them. The natural variability of the Amazon Forest provides an abundance of species of different sizes and shapes, conferring several peculiarities, which makes it necessary to use up-to-date and precise methods for timber quantification in Amazon Forest management. Given the employment of insufficient estimation methods for wood volume, this study scrutinizes the disparities between the actual harvested merchantable wood volume and the volume estimated by the forest inventory during the harvesting phase across five distinct public forest areas operating under sustainable forest management concessions. We used mixed-effect models to evaluate the relationships between inventory and harvested volume for genera and forest regions. We performed an equivalence test to assess the similarity between the volumes obtained during the pre-and post-harvest phases. We calculated root mean square error and percentage bias for merchantable volume as accuracy metrics. There was a strong tendency for the 100% forest inventory to overestimate merchantable wood volume, regardless of genus and managed area. There was a significant discrepancy between the volumes inventoried and harvested in different regions intended for sustainable forest management, in which only 22% of the groups evaluated were equivalent. The methods currently practiced by forest companies for determining pre-harvest merchantable volume are inaccurate enough to support sustainable forest management in the Amazon. They may even facilitate the region's illegal timber extraction and organized crime.

Current practices of conducting forest management plans in the Amazon may risk the survival of timber species

Measurement of the recovery of wood stocks in logging areas in the Amazon helps to understand the efficiency of policies for the use and conservation of native forests. This work evaluated the effects of logging on the dynamics and production of commercial species in the short and medium term in a conservation unit in the state of Rondônia. Structural patterns of the species as a function of mortality and recruitment rates, patterns of average increment in diameter, and estimates of forest production in the short and medium term were analyzed. The study was carried out in an area designated as Annual Production Unit 2 of Forest Management Unit III of Jamari National Forest. In addition to legalized harvesting, there are reports of illegal logging in the area as of 2015. Inventory data from 2011, 2015, and 2018 were used, considering trees of commercial value with a diameter at breast height (DBH) greater than 10 cm. Mortality rate, recruitment, periodic annual increment, absolute tree density, basal area, and commercial volume, by species and DBH classes, as well as the similarity of the species regarding the growth pattern. The population structure of species over the years was affected by tree mortality, mainly due to damage caused by illegal logging. Mean increment values varied by species and diameter classes, and six species represented 72% of the total volume of wood stock. It is important to review the criteria for sustainable forest production in the long term. Thus, it is necessary to promote species diversity and improve public authorities' capacity to enforce and of the private sector to comply with legislation. This, in turn, will enable the development of strategies to make the consumption of legal wood more rational.

Artificial intelligence in timber forensics employing DNA barcode database

Extreme difficulties in species identification of illegally sourced wood with conventional tools have accelerated illicit logging activities, leading to the destruction of natural resources in India. In this regard, the study primarily focused on developing a DNA barcode database for 41 commercial timber tree species which are highly vulnerable to adulteration in south India. The developed DNA barcode database was validated using an integrated approach involving wood anatomical features of traded wood samples collected from south India. Traded wood samples were primarily identified using wood anatomical features using IAWA list of microscopic features for hardwood identification. Consortium of Barcode of Life (CBOL) recommended barcode gene regions (rbcL, matK & psbA-trnH) were employed for developing DNA barcode database. Secondly, we employed artificial intelligence (AI) analytical platform, Waikato Environment for Knowledge Analysis (WEKA) for analyzing DNA barcode sequence database which could append precision, speed, and accuracy for the entire identification process. Among the four classification algorithms implemented in the machine learning algorithm (WEKA), best performance was shown by SMO, which could clearly allocate individual samples to their respective sequence database of biological reference materials (BRM) with 100 % accuracy, indicating its efficiency in authenticating the traded timber species. Major advantage of AI is the ability to analyze huge data sets with more precision and also provides a large platform for rapid authentication of species, which subsequently reduces human labor and time.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13205-023-03604-0.

Briquette production as a sustainable alternative for waste management in the tannin extraction industry

Tannins are polyphenols that can be extracted from different parts of the plant and have different known commercial applications. The extraction of tannins generates a significant amount of low-density wastes. An alternative that can be applied to ensure energy reuse of this low-density waste is a densification process, such as briquetting. Thus, this research work aimed to assess the energy potential of the residue from the extraction of tannins of different forest species, aiming at its energy reuse in the form of briquettes. Stem barks of 6 forest species (Acacia mangium, AM; Anacardium occidentale, AO; Anadenanthera colubrina, AC; Azadirachta indica, AI; Mimosa caesalpiniaefolia, MC; Mimosa tenuiflora, MT) were used after tannin extraction. Bulk density, energy (higher, lower, and net heating value) and thermal (TG/DTG) properties, and chemical constitution (extractives, lignin, and holocellulose) of the materials were determined. In addition, briquette compaction ratio, apparent density, volumetric expansion, and water adsorption over the days were evaluated. The highest bulk densities were found in the barks of Anadenanthera colubrina (0.529 g.cm^-3) and Mimosa tenuiflora (0.407 g.cm^-3), whereas the species that showed the best result of higher heating value was Acacia mangium (20.44 MJ.kg^-1), followed by Azadirachta indica (19.39 MJ.kg^-1) and Mimosa caesalpiniifolia (18.85 MJ.kg^-1). Briquetting increased the density of the evaluated material by 2.3 to 4.9 times. All briquettes produced with wastes from tannin extraction evaluated in this work showed potential for energy production. With more information on the quantification of waste generated and data on the economic viability of production, these tannin-producing industries can benefit both environmentally and economically, by reusing these wastes for energy production.

The drivers and impacts of Amazon forest degradation

Approximately 2.5 × 10⁶ square kilometers of the Amazon forest are currently degraded by fire, edge effects, timber extraction, and/or extreme drought, representing 38% of all remaining forests in the region. Carbon emissions from this degradation total up to 0.2 petagrams of carbon per year (Pg C year^-1), which is equivalent to, if not greater than, the emissions from Amazon deforestation (0.06 to 0.21 Pg C year^-1). Amazon forest degradation can reduce dry-season evapotranspiration by up to 34% and cause as much biodiversity loss as deforestation in human-modified landscapes, generating uneven socioeconomic burdens, mainly to forest dwellers. Projections indicate that degradation will remain a dominant source of carbon emissions independent of deforestation rates. Policies to tackle degradation should be integrated with efforts to curb deforestation and complemented with innovative measures addressing the disturbances that degrade the Amazon forest.

Some fundamental elements for studying social-ecological co-existence in forest common pool resources

For millennia, societies have tried to find ways to sustain people's livelihoods by setting rules to equitably and sustainably access, harvest and manage common pools of resources (CPR) that are productive and rich in species. But what are the elements that explain historical successes and failures? Elinor Ostrom suggested that it depends on at least eight axiomatic principles of good governance, whereas empirical results suggest that these principles are not sufficient to describe them, especially when applied to CPRs that possess great social and ecological diversity. The aim of this article is to explore the behavior of a mathematical model of multi-species forest dynamics that respects the foundations of ecology and Ostrom's governance theory, in order to detect possible constraints inherent to the functioning of these complex systems. The model reveals that fundamental structural laws of compatibilities between species life-history traits are indeed constraining the level of co-existence (average and variance) between a diversity of co-vulnerable timber resource users (RU) and of competing tree species. These structural constraints can also lead to unexpected outcomes. For instance in wetter forest commons, opening up the access to as many diverse RUs as there are competing tree species, produces a diversity of independently-controlled disturbances on species, collectively improving the chances of coexistence between species with different life-history traits. Similar benefits are observed on forest carbon and on profits from timber harvesting. However in drier forest commons, the same benefits cannot be observed, as predicted on the basis of the constraining laws. The results show that the successes and failures of certain management strategies can be reasonably explained by simple mechanistic theories from ecology and the social-ecological sciences, which are themselves constrained by fundamental ecological invariants. If corroborated, the results could be used, in conjunction with Ostrom's CPR theory, to understand and solve various human-nature coexistence dilemmas in complex social-ecological systems.

Clay and soil organic matter drive wood multi-elemental composition of a tropical tree species: Implications for timber tracing

Forensic methods to independently trace timber origin are essential to combat illegal timber trade. Tracing product origin by analysing their multi-element composition has been successfully applied in several commodities, but its potential for timber is not yet known. To evaluate this potential the drivers of wood multi-elemental composition need to be studied. Here we report on the first study relating wood multi-elemental composition of forest trees to soil chemical and physical properties. We studied the reactive soil element pools and the multi-elemental composition in sapwood and heartwood for 37 Azobé (Lophira alata) trees at two forest sites in Cameroon. A total of 46 elements were measured using ICP-MS. We also measured three potential drivers of soil and wood elemental composition: clay content, soil organic matter and pH. We tested associations between soil and wood using multiple regressions and multivariate analyses (Mantel test, db-RDA). Finally, we performed a Random Forest analysis of heartwood elemental composition to check site assignment accuracy. We found elemental compositions of soil, sapwood and heartwood to be significantly associated. Soil clay content and organic matter positively influenced individual element concentrations (for 13 and 9 elements out of 46 respectively) as well as the multi-elemental composition in wood. However, associations between wood and topsoil elemental concentrations were only significant for one element. We found close associations between element concentrations and composition in sapwood and heartwood. Lastly, the Random Forest assignment success was 97.3 %. Our findings indicate that wood elemental composition is associated with that in the topsoil and its variation is related to soil clay and organic matter content. These associations suggests that the multi-elemental composition of wood can yield chemical fingerprints obtained from sites that differ in soil properties. This finding in addition to the high assignment accuracy shows potential of multi-element analysis for tracing wood origin.

Knowledge Mapping on Nepal’s Protected Areas Using CiteSpace and VOSviewer

Protected areas (PAs) play a vital role in environmental conservation, particularly in Asian countries. Numerous studies were conducted on PAs in Nepal. We analyzed 864 papers from the Web of Science database using two visualization tools: VOSviewer and CiteSpace. This study identified the most influential journals, institutions, countries, and regions. In addition, we investigated the changing trend of research hotspots on PAs in Nepal. Keyword mapping was conducted for each type of PA and their differences were compared. We found that the research hotspots are changing with the shifting of conservation policies in Nepal. We suggest conducting more predictive studies on the future development of PAs. Currently, PA research is mainly conducted in traditional disciplines, but with the impact of climate change and the consequent increase in its negative impacts, academic contributions from other disciplines are expected to increase much more. We found that there was a shift in research power in countries and regions. We also detected an imbalanced distribution in which “protected areas” and “national parks” have been studied the most. Only 12 publications were about the hunting reserve, despite its importance to snow leopard conservation and economic significance to the buffer zone communities.

Investigating illegal activities that affect biodiversity: the case of wildlife consumption in the Brazilian Amazon

The illegal use of natural resources, manifested in activities like illegal logging, poaching, and illegal wildlife trade, poses a global threat to biodiversity. Addressing them will require an understanding of the magnitude of and factors influencing these activities. However, assessing such behaviors is challenging because of their illegal nature, making participants less willing to admit engaging in them. We compared how indirect (randomized response technique) and direct questioning techniques performed when assessing non-sensitive (fish consumption, used as negative control) and sensitive (illegal consumption of wild animals) behaviors across an urban gradient (small towns, large towns, and the large city of Manaus) in the Brazilian Amazon. We conducted 1,366 surveys of randomly selected households to assess the magnitude of consumption of meat from wild animals (i.e., wild meat) and its socioeconomic drivers, which included years the head of household lived in urban areas, age of the head of household, household size, presence of children, and poverty. The indirect method revealed higher rates of wildlife consumption in larger towns than did the direct method. Results for small towns were similar between the two methods. The indirect method also revealed socioeconomic factors influencing wild meat consumption that were not detected with direct methods. For instance, the indirect method showed that wild meat consumption increased with age of the head of household, and decreased with poverty and years the head of household lived in urban areas. Simultaneously, when responding to direct questioning, households with characteristics associated with higher wild meat consumption, as estimated from indirect questioning, tended to underreport consumption to a larger degree than households with lower wild meat consumption. Results for fish consumption, used as negative control, were similar for both methods. Our findings suggest that people edit their answers to varying degrees when responding to direct questioning, potentially biasing conclusions, and indirect methods can improve researchers' ability to identify patterns of illegal activities when the sensitivity of such activities varies across spatial (e.g., urban gradient) or social (e.g., as a function of age) contexts. This work is broadly applicable to other geographical regions and disciplines that deal with sensitive human behaviors.

Application of remote sensing in environmental impact assessment: a case study of dam rupture in Brumadinho, Minas Gerais, Brazil

The collapse of mining tailing dams in Brumadinho, Minas Gerais, Brazil, that occurred in 2019 was one of the worst environmental and social disasters witnessed in the country. In this sense, monitoring any impacted areas both before and after the disaster is crucial to understand the actual scenario and problems of disaster management and environmental impact assessment. In order to find answers to that problem, the aim of this study was to identify and analyze the spatiality of the impacted area by rupture of the tailing dam of the Córrego do Feijão mine in Brumadinho, Minas Gerais, by using orbital remote sensing. Land use and land occupation, phytoplankton chlorophyll-a, water turbidity, total suspended solids on water, and carbon sequestration efficiency by vegetation (CO₂Flux) were estimated by orbital imagery from the Landsat-8/OLI and MSI/Sentinel-2 sensors in order to assess the environmental impacts generated by the disaster. Data were extracted from spectral models in which the variables that best demonstrated the land use variation over the years were sought. Mean comparison by t-test was performed to compare the time series analyzed, that is, before and after the disaster. Through the analysis of water quality, it was observed that the environmental impact was calamitous to natural resources, especially water from Córrego do Feijão.

Tracking the impacts of El Niño drought and fire in human-modified Amazonian forests

With humanity facing an unprecedented climate crisis, the conservation of tropical forests has never been so important - their vast terrestrial carbon stocks can be turned into emissions by climatic and human disturbances. However, the duration of these effects is poorly understood, and it is unclear whether impacts are amplified in forests with a history of previous human disturbance. Here, we focus on the Amazonian epicenter of the 2015-16 El Niño, a region that encompasses 1.2% of the Brazilian Amazon. We quantify, at high temporal resolution, the impacts of an extreme El Niño (EN) drought and extensive forest fires on plant mortality and carbon loss in undisturbed and human-modified forests. Mortality remained higher than pre-El Niño levels for 36 mo in EN-drought-affected forests and for 30 mo in EN-fire-affected forests. In EN-fire-affected forests, human disturbance significantly increased plant mortality. Our investigation of the ecological and physiological predictors of tree mortality showed that trees with lower wood density, bark thickness and leaf nitrogen content, as well as those that experienced greater fire intensity, were more vulnerable. Across the region, the 2015-16 El Niño led to the death of an estimated 2.5 ± 0.3 billion stems, resulting in emissions of 495 ± 94 Tg CO2 Three years after the El Niño, plant growth and recruitment had offset only 37% of emissions. Our results show that limiting forest disturbance will not only help maintain carbon stocks, but will also maximize the resistance of Amazonian forests if fires do occur.

Assessing invertebrate herbivory in human-modified tropical forest canopies

Studies on the effects of human-driven forest disturbance usually focus on either biodiversity or carbon dynamics but much less is known about ecosystem processes that span different trophic levels. Herbivory is a fundamental ecological process for ecosystem functioning, but it remains poorly quantified in human-modified tropical rainforests.Here, we present the results of the largest study to date on the impacts of human disturbances on herbivory. We quantified the incidence (percentage of leaves affected) and severity (the percentage of leaf area lost) of canopy insect herbivory caused by chewers, miners, and gall makers in leaves from 1,076 trees distributed across 20 undisturbed and human-modified forest plots in the Amazon.We found that chewers dominated herbivory incidence, yet were not a good predictor of the other forms of herbivory at either the stem or plot level. Chewing severity was higher in both logged and logged-and-burned primary forests when compared to undisturbed forests. We found no difference in herbivory severity between undisturbed primary forests and secondary forests. Despite evidence at the stem level, neither plot-level incidence nor severity of the three forms of herbivory responded to disturbance. Synthesis. Our large-scale study of canopy herbivory confirms that chewers dominate the herbivory signal in tropical forests, but that their influence on leaf area lost cannot predict the incidence or severity of other forms. We found only limited evidence suggesting that human disturbance affects the severity of leaf herbivory, with higher values in logged and logged-and-burned forests than undisturbed and secondary forests. Additionally, we found no effect of human disturbance on the incidence of leaf herbivory.

Computer vision-based wood identification and its expansion and contribution potentials in wood science: A review

The remarkable developments in computer vision and machine learning have changed the methodologies of many scientific disciplines. They have also created a new research field in wood science called computer vision-based wood identification, which is making steady progress towards the goal of building automated wood identification systems to meet the needs of the wood industry and market. Nevertheless, computer vision-based wood identification is still only a small area in wood science and is still unfamiliar to many wood anatomists. To familiarize wood scientists with the artificial intelligence-assisted wood anatomy and engineering methods, we have reviewed the published mainstream studies that used or developed machine learning procedures. This review could help researchers understand computer vision and machine learning techniques for wood identification and choose appropriate techniques or strategies for their study objectives in wood science.

Environmental Threats over Amazonian Indigenous Lands

This study investigates the main threats related to environmental degradation that affect Amazonian Indigenous Lands (ILs). Through a cluster analysis, we group ILs according to the set of common environmental threats that occur within and outside their limits. The results show that most of the 383 ILs are affected internally by a combination of different environmental threats, namely: deforestation, forest degradation, fires, mining, croplands, pastures, and roads. However, the ILs affected by multiple and relatively severe threats are mainly located in the arc of deforestation and the Roraima state. The threats related to forest loss (deforestation, forest degradation, and fires) are more intense in the ILs’ buffer zones than within, showing that ILs effectively promote environmental preservation. In the cluster analysis, we identified seven clusters that are characterized by common environmental threats within and around their limits, and, based on these results, we have outlined four environmental policy priorities to be strengthened and applied in Amazonian ILs: protecting ILs’ buffer zones; strengthening surveillance actions, and combating illegal deforestation, forest degradation, and mining activities in ILs; preventing and fighting fires; and removing invaders from all ILs in the Amazon. In this study, we warn that the threats presented make the Indigenous peoples in the Amazon more vulnerable. To guarantee indigenous peoples’ rights, illegal actions in these territories and their surroundings must be contained, and quickly.

Genome assemblies for two Neotropical trees: Jacaranda copaia and Handroanthus guayacan

The lack of genomic resources for tropical canopy trees is impeding several research avenues in tropical forest biology. We present genome assemblies for two Neotropical hardwood species, Jacaranda copaia and Handroanthus (formerly Tabebuia) guayacan, that are model systems for research on tropical tree demography and flowering phenology. For each species, we combined Illumina short-read data with in vitro proximity-ligation (Chicago) libraries to generate an assembly. For Jacaranda copaia, we obtained 104X physical coverage and produced an assembly with N50/N90 scaffold lengths of 1.020/0.277 Mbp. For H. guayacan, we obtained 129X coverage and produced an assembly with N50/N90 scaffold lengths of 0.795/0.165 Mbp. J. copaia and H. guayacan assemblies contained 95.8% and 87.9% of benchmarking orthologs, although they constituted only 77.1% and 66.7% of the estimated genome sizes of 799 and 512 Mbp, respectively. These differences were potentially due to high repetitive sequence content (>59.31% and 45.59%) and high heterozygosity (0.5% and 0.8%) in each species. Finally, we compared each new assembly to a previously sequenced genome for Handroanthus impetiginosus using whole-genome alignment. This analysis indicated extensive gene duplication in H. impetiginosus since its divergence from H. guayacan.

Deforestation Trajectories on a Development Frontier in the Brazilian Amazon: 35 Years of Settlement Colonization, Policy and Economic Shifts, and Land Accumulation

We examine deforestation processes in Apuí, a deforestation hotspot in Brazil's state of Amazonas and present processes of land-use change on this Amazonian development frontier. Settlement projects attract agents whose clearing reflects land accumulation and the economic importance of deforestation. We used a mixed-method approach in the Rio Juma Settlement to examine colonization and deforestation trajectories for 35 years at three scales of analysis: the entire landscape, cohorts of settlement lots divided by occupation periods, and lots grouped by landholding size per household. All sizes of landholdings are deforesting much more than before, and current political and economic forces favoring the agribusiness sector foreshadow increasing rates of forest clearing for pasture establishment in Apuí. The area cleared per year over the 2013-2018 period in Apuí grew by a percentage more than twice the corresponding percentage for the Brazilian Amazon as a whole. With the national congress and presidential administration signaling impunity for illegal deforestation, wealthy actors, and groups are investing resources in land grabbing and land accumulation, with land speculation being a crucial deforestation factor. This paper is unique in providing causal explanations at the decision-maker's level on how deforestation trajectories are linked to economic and political events (period effects) at the larger scales, adding to the literature by showing that such effects were more important than aging and cohort effects as explanations for deforestation trajectories. Additional research is needed to deepen our understanding of relations between land speculation, illegal possession of public lands, and the expansion of agricultural frontiers in Amazonia.

Assessment of the Local Perceptions on the Drivers of Deforestation and Forest Degradation, Agents of Drivers, and Appropriate Activities in Cambodia

Understanding the drivers of deforestation and forest degradation and the agents of such drivers is important for introducing appropriate policy interventions. Here, we identified drivers and agents of drivers through the analysis of local perceptions using questionnaire surveys, focus group discussions, and field observations. The Likert scale technique was employed for designing the questionnaire with scores ranging from 1 (strongly disagree) to 5 (strongly agree). We found nine direct drivers of forest deforestation and forest degradation, namely illegal logging (4.53 ± 0.60, ± is for standard deviation), commercial wood production (4.20 ± 0.71), land clearing for commercial agriculture (4.19 ± 1.15), charcoal production (3.60 ± 1.12), land clearing for subsistence agriculture (3.54 ± 0.75), new settlement and land migration (3.43 ± 0.81), natural disasters (3.31 ± 0.96), human-induced forest fires (3.25 ± 0.96), and fuelwood for domestic consumption (3.21 ± 0.77). We also found four main indirect drivers, namely lack of law enforcement, demand for timber, land tenure right, and population growth. Our analysis indicates that wood furniture makers, medium and large-scale agricultural investors, charcoal makers, land migrants, firewood collectors, and subsistent farmers were the agents of these drivers. Through focus group discussions, 12 activities were agreed upon and could be introduced to reduce these drivers. In addition to enforcing the laws, creating income-generating opportunities for locals along with the provision of environmental education could ensure long-term reduction of these drivers. The REDD+ project could be an option for creating local income opportunities, while reducing deforestation and forest degradation.

Stocks of Carbon in Logs and Timber Products from Forest Management in the Southwestern Amazon

Amazon forest management plans have a variety of effects on carbon emissions, both positive and negative. All of these effects need to be quantified to assess the role of this land use in climate change. Here, we contribute to this effort by evaluating the carbon stocks in logs and timber products from an area under forest management in the southeastern portion of Acre State, Brazil. One hundred and thirty-six trees of 12 species had DBH ranging from 50.9 cm to 149.9 cm. Basic wood density ranged from 0.3 cm−3 to 0.8 g cm−3 with an average of 0.6 g cm−3. The logs had a total volume of 925.2 m3, biomass of 564 Mg, and carbon stock of 484.2 MgC. The average volumetric yield coefficient (VYC) was 52.3% and the carbon yield coefficient (CYC) was 53.2% for logs of the 12 species. The sawn-wood products had a total volume of 484.2 m3, biomass of 302.6 Mg, and carbon stock of 149.9 MgC. Contributions of the different species to the total carbon stored in sawn-wood products ranged from 2.2% to 21.0%. Means and standard deviations for carbon transferred to sawn-wood products per-species from the 1252.8-ha harvested area ranged from 0.4 ± 1.1 MgC to 2.9 ± 0.4 MgC, with the largest percentages of the total carbon stored in wood products being from Dipteryx odorata (21.0%), Apuleia leiocarpa (18.7%), and Eschweilera grandiflora (11.7%). A total of 44,783 pieces of sawn lumber (such as rafters, planks, boards, battens, beams, and small beams) was obtained from logs derived from these trees. Lumber production was highest for boards (54.6% of volume, 47.4% of carbon) and lowest for small beams (1.9% of volume, 2.3% of carbon). The conversion factor for transforming log volume into carbon stored in sawn-wood products was 16.2%. Our results also show that species that retain low amounts of carbon should be allowed to remain in the forest, thereby avoiding low sawmill yield (and consequent generation of waste) and allowing these trees to continue fulfilling environmental functions.

Machine Learning Models with Quantitative Wood Anatomy Data Can Discriminate between Swietenia macrophylla and Swietenia mahagoni

Illegal logging and associated trade aggravate the over-exploitation of Swietenia species, of which S. macrophylla King, S. mahagoni (L.) Jacq, and S. humilis Zucc. have been listed in Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) Appendix Ⅱ. Implementation of CITES necessitates the development of efficient forensic tools to identify wood species accurately, and ideally ones readily deployable in wood anatomy laboratories across the world. Herein, a method using quantitative wood anatomy data in combination with machine learning models to discriminate between three Swietenia species is presented, in addition to a second model focusing only on the two historically more important species S. mahagoni and S. macrophylla. The intra- and inter-specific variations in nine quantitative wood anatomical characters were measured and calculated based on 278 wood specimens, and four machine learning classifiers—Decision Tree C5.0, Naïve Bayes (NB), Support Vector Machine (SVM), and Artificial Neural Network (ANN)—were used to discriminate between the species. Among these species, S. macrophylla exhibited the largest intraspecific variation, and all three species showed at least partly overlapping values for all nine characters. SVM performed the best of all the classifiers, with an overall accuracy of 91.4% and a per-species correct identification rate of 66.7%, 95.0%, and 80.0% for S. humilis, S. macrophylla, and S. mahagoni, respectively. The two-species model discriminated between S. macrophylla and S. mahagoni with accuracies of over 90.0% using SVM. These accuracies are lower than perfect forensic certainty but nonetheless demonstrate that quantitative wood anatomy data in combination with machine learning models can be applied as an efficient tool to discriminate anatomically between similar species in the wood anatomy laboratory. It is probable that a range of previously anatomically inseparable species may become identifiable by incorporating in-depth analysis of quantitative characters and appropriate statistical classifiers.

Machine learning approaches outperform distance- and tree-based methods for DNA barcoding of Pterocarpus wood

Machine-learning approaches (MLAs) for DNA barcoding outperform distance- and tree-based methods on identification accuracy and cost-effectiveness to arrive at species-level identification of wood. DNA barcoding is a promising tool to combat illegal logging and associated trade, and the development of reliable and efficient analytical methods is essential for its extensive application in the trade of wood and in the forensics of natural materials more broadly. In this study, 120 DNA sequences of four barcodes (ITS2, matK, ndhF-rpl32, and rbcL) generated in our previous study and 85 downloaded from National Center for Biotechnology Information (NCBI) were collected to establish a reference data set for six commercial Pterocarpus woods. MLAs (BLOG, BP-neural network, SMO and J48) were compared with distance- (TaxonDNA) and tree-based (NJ tree) methods based on identification accuracy and cost-effectiveness across these six species, and also were applied to discriminate the CITES-listed species Pterocarpus santalinus from its anatomically similar species P. tinctorius for forensic identification. MLAs provided higher identification accuracy (30.8-100%) than distance- (15.1-97.4%) and tree-based methods (11.1-87.5%), with SMO performing the best among the machine learning classifiers. The two-locus combination ITS2 + matK when using SMO classifier exhibited the highest resolution (100%) with the fewest barcodes for discriminating the six Pterocarpus species. The CITES-listed species P. santalinus was discriminated successfully from P. tinctorius using MLAs with a single barcode, ndhF-rpl32. This study shows that MLAs provided higher identification accuracy and cost-effectiveness for forensic application over other analytical methods in DNA barcoding of Pterocarpus wood.

Illegal Selective Logging and Forest Fires in the Northern Brazilian Amazon

Illegal selective logging and forest fires occur on a large scale in the northern Brazilian Amazon, contributing to an increase in tree mortality and a reduction in forest carbon stock. A total of 120 plots of 0.25 ha (30 ha) were installed in transitional ecosystems or ecotones (LOt) between the forested shade-loving campinarana (Ld) and dense-canopy rainforest, submontane (Ds), in the National Forest (Flona) of Anauá, southern Roraima. Measuring the diameters at breast height (DBH ≥ 10 cm) and the heights of 171 dead trees (fallen naturally, illegally exploited, and affected by forest fires), enabled the estimation of carbon content from the application of a biomass equation developed at Manaus, and the calculation of a correction factor, using the average height of the largest trees. From 2015–2017, we mapped the real extent of illegal selective logging and forest fires across the region with CLASlite and INPE/Queimadas. From measurements of 14,730 live and dead trees across 30 hectares (491 ± 15 trees·ha−1), the illegal selective logging and associated forest fires, and aggravation by severe El Niño droughts resulted in an 8.2% mortality of trees (40 ± 9 dead trees·ha−1) and a 3.5% reduction in forest carbon stock (6 ± 3 Mg·ha−1) in the short-term. The surface area or influence of forest fires of very high density were estimated in the south-central region of Roraima (8374 km²) and the eastern region of the Flona Anauá (37 km²). Illegal selective logging and forest fires in forest areas totaled 357 km² in the mosaic area, and 6 km² within Flona Anaua. Illegal selective logging and forest fires in the years of severe El Niño droughts threatened the maintenance of environmental services provided by Amazonian forests.