Accurate estimation of the volume and above-ground biomass of exploitable trees by the practice of selective logging is essential for the elaboration of a sustainable management plan. The objective of this study is to develop machine learning models capable of estimating the volume and biomass of commercial trees in the Southwestern Amazon, based on dendrometric, climatic and topographic characteristics. The study was carried out in the municipality of Porto Acre, Acre state, Brazil. The volume and biomass of sample trees were determined using dendrometric, climatic and topographic variables. The Boruta algorithm was applied to select the best set of variables. Support Vector Machines (SVM), Artificial Neural Networks (ANN), Random Forests (RF) and the Generalized Linear Model (GLM) were the machine learning methods evaluated. In general, the evaluated methods showed a satisfactory generalization power. The results showed that the volume and biomass predictions of commercial trees in the Amazon rainforest differed between the techniques (p < 0.05). ANNs showed the best performance in predicting the volume and biomass of commercial trees, with the highest r_yŷ and the lowest RSME and MAE. Thus, machine learning methods such as SVM, ANN, RF and GLM are shown to be useful and efficient tools for estimating the volume and biomass of commercial trees in the Amazon rainforest. These methods can be useful tools to improve the accuracy of estimates in forest management plans.

Necromass has a relevant role to play in the carbon stock of forest ecosystems, especially with the increase of tree mortality due to climate change. Despite this importance, its quantification is often neglected in tropical forests. The objective of this study was to quantify the carbon storage in a secondary Atlantic Forest fragment in Viçosa, Minas Gerais, Brazil. Coarse Woody Debris (CWD), standing dead trees (snags), and litter were quantified in twenty 10 m x 50 m plots randomly positioned throughout the forest area (simple random sampling). Data were collected during 2015, from July to December. The CWD and snags volumes were determined by the Smalian method and by allometric equations, respectively. The necromass of these components was estimated by multiplying the volume by the apparent density at each decomposition classes. The litter necromass was estimated by the proportionality method and the average of the extrapolated estimates per hectare. The carbon stock of the three components was quantified by multiplying the necromass and the carbon wood content. The total volume of dead wood, including CWD and snag, was 23.6 ± 0.9 m³ ha⁻¹, being produced mainly by the competition for resources, senescence, and anthropic and climatic disturbances. The total necromass was 16.3 ± 0.4 Mg ha⁻¹. The total carbon stock in necromass was 7.3 ± 0.2 MgC ha⁻¹. The CWD, snag and litter stocked 3.0 ± 0.1, 1.8 ± 0.1, and 2.5 ± 0.1 MgC ha⁻¹, respectively. These results demonstrate that although necromass has a lower carbon stock compared to biomass, neglecting its quantification may lead to underestimation of the carbon balance of forest ecosystems and their potential to mitigate climate change.

The grapple skidder is a self-propelled forestry machine that is used for the extraction of trees in wood harvesting—commonly used in full tree systems. Moving this machine can expose operators to occupational hazards of physical origin, among which whole body vibration stands out. However, the measurement of this risk agent does not have a standard measurement time, being performed for periods of approximately 30 min—disregarding the time of the daily workday. In view of this, it was analyzed whether occupational exposure to whole body vibration transmitted to grapple skidder operators using different measurement times complies with the guidelines for preventative purposes. Thus, measurements of whole-body vibration were carried out along three orthogonal axes over a period of one hour and over a daily workday of eight hours—by which were measured the daily (8 h) vibration exposure for the l-axis and the vibration value. The acceleration values in the three evaluated axes were higher for the daily working day, denoting the influence of the measurement time. In addition, the vibration dose value resulted in values above the action limit for both evaluations; however, the daily workload was highlighted—indicating the presence of higher vibration peaks over a longer measurement time. Thus, the assertiveness and influence of measurement times over the daily working day for whole-body vibration transmitted to grapple skidder operators is evidenced.

The potential of tree component neutralization was studied to understand the relationship between livestock, forest production and greenhouse gas (GHG) emissions. The objective of this work was to evaluate the carbon balance in a silvopastoral system including semi-intensive dairy farming and to quantify emissions per unit milk produced in Visconde do Rio Branco, Minas Gerais state, Brazil. The tree component of the silvopastoral system was a hybrid of Eucalyptus grandis x Eucalyptus urophylla in a semi-intensive production system, including Brachiaria decumbes as forage grass and crossbred dairy cow. Carbon stock was calculated from the estimated wood volume with forest inventory data, wood density and wood carbon content factor for tree species. Livestock (enteric fermentation and animal waste management), electricity, fuel consumption, nitrogen and limestone fertilization were the GHG emissions considered using emission factors. The carbon balance was obtained by comparing annually the GHG emissions with the increase of this element. The Average Annual Carbon Increment (IMAC) was 7.631 Mg CO_2e ha⁻¹year⁻¹ and the Annual Average Emission of 5.817 Mg CO_2e ha⁻¹year⁻¹. The carbon balance, in all the years, was positive evidencing the contribution of the silvopastoral system to minimize GHG emissions and neutralize the emissions in milk production.

Predicting wood biomass and carbon stock contents in planted forests can vary due to limitations associated with the measurement of parameters. Therefore, reducing possible errors generated over biomass and carbon stock quantification is an important step in obtaining reliable data. The study aimed to compare the use of destructive and non-destructive methodologies for predicting biomass and carbon stock in a planted Eucalyptus forest. Scaling was performed on 21 trees and 3 methodologies for carbon stock estimation were compared. For methodology 1, a control sample was harvested, sectioned, weighted in the field, and the carbon stock calculated based on these data. Methodology 2 was also destructive, as trees were harvested, scaled and the carbon stock predicted based on these data. Methodology 3 was non-destructive, as trees were scaled upright with the aid of equipment and the predicted carbon stock was based on these data. Biomass and carbon stock were compared by Test F and no statistical difference was observed. The data were separated according to diametric classes and compared by the Kolmogorov-Smirnov test, and again no significant difference was observed. Furthermore, three equations were generated based on the Schumacher & Hall model and compared by the identity test model and no differences between the methodologies were observed. Thus, both nondestructive and destructive methodologies herein evaluated were effective and showed equal results to the control sample. Moreover, the use of the non-destructive methodology reduces time and cost destined to predicting biomass and carbon stock.