This research aimed to develop statistical models to predict basal area increment (BAI) for Araucaria angustifolia using Artificial Neural Networks (ANNs). Tree species were measured for their biometric variables and identified at the species level. The data were subdivided into three groups: (1) intraspecific competition with A. angustifolia; (2) the first group of species that causes interspecific competition with A. angustifolia; and (3) the second group of species that causes interspecific competition with A. angustifolia. We calculated both the dependent and independent distance and the described competition indices, considering the impact of group stratification. Multi-layer Perceptron (MLP) ANN was structured for modeling. The main results were that: (i) the input variables size and competition were the most significant, allowing us to explain up to 77% of the A. angustifolia BAI variations; (ii) the spatialization of the competing trees contributed significantly to the representation of the competitive status; (iii) the separate variables for each competition group improved the performance of the models; and (iv) besides the intraspecific competition, the interspecific competition also proved to be important to consider. The ANN developed showed precision and generalization, suggesting it could describe the increment of a species common in native forests in Southern Brazil and with potential for upcoming forest management initiatives.

Crown variables are used in growth and production models to increase the accuracy of estimates. These variables are costly and difficult to measure, mainly in dense native forests such as mixed omb...

Artificial intelligence (AI) seeks to simulate the human ability to reason, make decisions, and solve problems. Several AI methodologies have been introduced in forestry to reduce costs and increase accuracy in estimates. We evaluate the performance of Artificial Neural Networks (ANN) in estimating the heights of Araucaria angustifolia (Bertol.) Kuntze (Brazilian pine) trees. The trees are growing in Uneven-aged Mixed Forests (UMF) in southern Brazil and are under different levels of competition. The dataset was divided into training and validation sets. Multi-layer Perceptron (MLP) networks were trained under different Data Normalization (DN) procedures, Neurons in the Hidden Layer (NHL), and Activation Functions (AF). The continuous input variables were diameter at breast height (DBH) and height at the base of the crown (HCB). As a categorical input variable, we consider the sociological position of the trees (dominant–SP1 = 1; codominant–SP2 = 2; and dominated–SP3 = 3), and the continuous output variable was the height (h). In the hidden layer, the number of neurons varied from 3 to 9. Results show that there is no influence of DN in the ANN accuracy. However, the increase in NHL above a certain level caused the model’s over-fitting. In this regard, around 6 neurons stood out, combined with logistic sigmoid AF in the intermediate layer and identity AF in the output layer. Considering the best selected network, the following values of statistical criteria were obtained for the training dataset (R² = 0.84; RMSE = 1.36 m, and MAPE = 6.29) and for the validation dataset (R² = 0.80; RMSE = 1.49 m, and MAPE = 6.53). The possibility of using categorical and numerical variables in the same modeling has been motivating the use of AI techniques in different forestry applications. The ANN presented generalization and consistency regarding biological realism. Therefore, we recommend caution when determining DN, amount of NHL, and using AF during modeling. We argue that such techniques show great potential for forest management procedures and are suggested in other similar environments.

Equations express that height-diameter relationships are used to estimate tree heights that were not measured in the plots, as well as to calculate their volumes. In this study, we modelled height for Eucalyptus grandis Hill ex. Maiden stands using nonlinear mixed effect models in Mocuba District, Central Mozambique. Models were tested from 1414 trees measured, in 40 plots with dimensions of 20 x 20 m. Model one (M1) was the best in relation to the others according to the evaluated criteria (R2 = 0.9183; RMSE = 0.558; AIC = -1234.8; BIC = -1224.9). The inclusion of the variables dominant height (h100, m), basal area (G, m2.ha-1) and mean basal area diameter (dg, m) in model 1 (M1) within the scope of mixed effect model structures that involved both fixed and random effect parameters (Model 7) provides better fitting and more precise predictions. For calibration of random effects according to plot, height measurement of at least five trees is required (four dominant trees + one tree randomly selected in plot). Our model is site-specific and its application should therefore be limited to the stands with characteristics that were the basis of this study. Further works on recalibration, validation, and verification of our model using a larger dataset collected from a wider range of species distribution will be more interesting.

This paper presents a simulation of the regulation of Araucaria angustifolia (Bertol.) Kuntze timber stocks using Liocourt’s law. Although this species is currently protected by law, recent government initiatives are being considered to propose sustainable forest management practices by selecting small rural properties in Southern Brazil. Here, we simulate the applicability of Liocourt’s law in a typical rural property, the size of which is approximately 85 ha. Forest inventory measurements were conducted by estimating the diameter at the breast height (d), total height (h), and annual diameter increments of 308 trees that fit the criteria of d ≥ 10 cm, distributed on 35 permanent plots of 400 m² each. As a result, a reverse J-shaped d distribution was found. On average, 303 trees can be found per hectare (ha). Local allometric equations showed their basal area (G) to be 21.9 m²∙ha⁻¹, and their commercial volume (V) to be 172 m³∙ha⁻¹, whereas Liocourt’s quotient (q) was 1.31. Based on these attributes, nine different forest management scenarios were proposed by simulating a remaining basal area (Grem) of 10.0, 14.0, and 18.0 m²∙ha⁻¹, and Liocourt’s quotient was changed to 1.1, 1.3, and 1.5. All scenarios consider a d of 62.5 cm. In the less intensive scenario (i.e., q value = 1.5 and larger basal area of 18.0 m²·ha⁻¹) there is greater optimization of space, and higher economic return is ensured to rural producers due to the definition of shorter cutting cycles. This also allows a faster growth rate in both d and h for smaller trees, due to the higher incidence of light onto the lower canopy layer, increasing the natural regeneration implementation of other native species. Forest management should thus be considered a goal in addition to consumer market characteristics for defining the ideal timber stock scenario.