Machine Learning for the Estimation of Diameter Increment in Mixed and Uneven-Aged Forests

Comparison between graph theory connectivity indices and landscape connectivity metrics for modeling river water quality in the southern Caspian sea basin

The maintenance of connectivity is critical to the proper functioning of an ecosystem. The present study was conducted with the aim of comparing graph theory connectivity indices and landscape connectivity metrics for the purpose of modeling river water quality. To conduct this study, a forest layer was extracted from land cover map and 25 large watersheds were selected. River water quality was then assessed from the perspective of 8 landscape connectivity metrics and 12 graph theory indices. We developed predictive models using stepwise linear regression, power, exponential, and logarithmic models to locate the best model form for each water quality parameter (dependent variable) we examined. The results indicated that models developed using graph theory connectivity indices resulted in higher coefficients of determination (R²) than models developed using landscape metrics. Only 5 independent variables from a potential set of 13 were significant in explaining the variation in water quality parameters. Also, the models with the highest R² attempted to explain variations in CO₃ (0.818), water discharge (0.733), and Ca levels (0.702). Therefore, the results of this study showed that graph theory connectivity indices had more significant correlation with water quality parameters compared to landscape connectivity metrics. This work also indicates that there exist nonlinear relationships among connectivity indices and water quality parameters.

Individual Tree Basal Area Increment Models for Brazilian Pine (Araucaria angustifolia) Using Artificial Neural Networks

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.

Projection Matrix Models: A Suitable Approach for Predicting Sustainable Growth in Uneven-Aged and Mixed Hyrcanian Forests

The Hyrcanian forests of Iran are mainly managed with the single-selection silvicultural technique. Despite significant ecological benefits associated with selection cutting, this type of forest management leads towards more challenging situations where it is difficult to maintain and practice successful forestry than in even-aged systems. Therefore, this study provides relevant management tools in the form of models to estimate low growth levels in Hyrcanian forests. In the present study, estimation of the population growth rate and then the allowable cut rate of these forests using a matrix model have been calculated in the Gorazbon district. For this purpose, the data of 256 permanent sample plots measured during the years between 2003 and 2012, as well as the data recorded about the trees harvested according to the forestry plan, have been used. As a first step, the most frequently occurring tree species were divided into four groups (beech, hornbeam, chestnut-leaved oak, and other species). Compartments of the district were divided into two groups of logged and unlogged compartments. The purpose of this division was to estimate the allowable cut and compare its volume with the volumes of observed and predicted allowable cuts obtained from forestry plans. The results showed that the total operated allowable cut (OAC) in logged compartments was more than the estimated allowable cut (EAC). In unlogged compartments, the total predicted allowable cut (PAC) was more than EAC. A comparison of EAC and OAC showed that hornbeam has been harvested more than its potential. However, chestnut-leaved oak and other species group have depicted opposite trends. Our models provide important advancements for estimating allowable cut that can enhance the goal of practicing sustainable forestry.