A Comprehensive Comparison of Artificial Neural Networks and Some Regression Modeling Techniques on Dominant Height Predictions of Oriental Spruce in a Mixed Forest, the Northeast Turkey

Background: Dominant height is needed for assessing silvicultural practices in sustainable wood production management. Also, dominant height is used as an important explanatory variable in forest growth and yield models. This paper introduces the evaluation for Artificial Neural Networks and Some Regression Modeling Techniques on Dominant Height Predictions of Oriental Spruce in a Mixed Forest, the Northeast Turkey. Methods: In this study, 873 height-age pairs were obtained from oriental spruce trees in a mixed forest stand. Nonlinear mixed-effects models (NLMEs), autoregressive models (ARM), dummy variable method (DVM), and artificial neural networks (ANNs) were compared to predict dominant height growth. Results: The best predictive model was NLME with single random parameter (root mean square error, RMSE: 0.68 m). The results showed that NLMEs outperformed ARM (RMSE: 1.09 m), DVM in conjunction with ARM (RMSE: 1.09 m), and ANNs (RMSE: from 1.11 to 2.40 m) in majority of the cases. Whereas considering variations among observations by random parameter(s) significantly improved predictions of dominant height, taking into account correlated error terms by autoregressive correlation parameter(s) enhanced slightly the predictions. ANNs generally underperformed compared to NLMEs, ARM, and DVM with ARM. Conclusion: All regression techniques fulfilled the desirable characteristics such as sigmoidal pattern, polymorphism, multiple asymptote, base-age invariance, and inflection point. However, ANNs could not most of these features excluding sigmoidal pattern. Accordingly, ANNs seem to insufficient to assure biological growth assumptions regarding dominant height growth.

GROWTH AND YIELD OF EVEN-AGED Cupressus lusitanica PLANTATIONS IN SOUTHERN BRAZIL

Plantation forestry in southern Brazil demands additional timber species to a higher market differentiation by providing high quality timber and exploitation of market niches. Cupressus lusitanica has long been recognized for this purpose but, until now, it was not properly region-wide quantified in terms of growth and yield. The present study delivers the lacking quantitative approach, which may encourage the commercial use of the species. With this study it was aimed at collecting and processing quantitative data from all known C. lusitanica stands in southern Brazil. Inventories were carried out (60 ha, 6-39 years of age) in order to model the development of dominant height (h100), basal area, volume and dominant diameter (d100). Dominant height was the basis for site quality evaluation, delivering site index curves, which, together with the commercial volume of the stands, allowed yield modelling. A wide amplitude of dominant height growth was detected (10-30 m at 20 years), indicating a great site quality variation. At age of 20 years, commercial volumes of 110 and 620 m³ ha-1 were observed for site indexes of 14 and 26, respectively, equivalent to a maximum of 6-31 m³ ha-1 year-1 at ages between 16-18 years. Results demonstrated in a robust manner that C. lusitanica has a high potential for cultivation in southern Brazil. Thus, offering the opportunity of market differentiation by promoting market niches whose demands timber for special solid end-uses.

Modeling Dominant Height Growth of Cedar (Cedrus libani A. Rich) Stands in Turkey

Sustainable forest management requires accurate prediction from a growth and yield system. Such a system relies heavily on some measure of site productivity, which is often the site index. A model was developed for predicting dominant height growth and site index of even-aged cedar (Cedrus libani A. Rich.) stands in Turkey. Stem-analysis data from 148 trees were used for model development and validation. Six dynamic height–age equations were derived using the generalized algebraic difference approach (GADA). Autocorrelation was modeled by expanding the error term as an autoregressive process. Based on numerical and graphical analysis, a GADA formulation derived from the Chapman–Richards model was selected. Based on relative error in dominant height prediction, 80 years was selected as the best reference age. The resulting equation provided the best compromise between biological and statistical aspects and, therefore, is recommended for height growth prediction and site classification of cedar stands in Turkey.

Dominant height growth and dynamic site index models for Crimean pine in the Kastamonu–Taşköprü region of Turkey

Some dynamic site index models based on the generalized algebraic difference approach (GADA) were fitted for Crimean pine (Pinus nigra J.F. Arnold subsp. pallasiana (Lamb.) Holmboe) stands in Taşköprü, Turkey. Data were obtained from 132 dominant trees representing the wide range of site quality in the region. Nonlinear regression analysis and a second-order continuous-time autoregressive error structure were applied. After autoregressive modeling, the fitted models were evaluated both statistically and graphically. The best results were obtained with the dynamic site index model derived from the Bertalanffy–Richards base equation, accounting for about the 99% of the total variance in height–age relationships in dominant trees, with an Akaike information criterion (AIC) value of 119.55 and root mean square error (RMSE) of 0.5446. The selected base-age invariant dynamic site index curves provided the polymorphism with multiple asymptotes and other realistic height growth patterns.

SITE QUALITY CURVES FOR AFRICAN MAHOGANY PLANTATIONS IN BRAZIL

ABSTRACT Site quality estimation is an important tool in forest management since it is useful for modeling growth and yield for even-aged stands. Data from African mahogany (Khaya ivorensis A. Chev.) Brazilian plantations were used to develop a model to predict dominant height growth, comparing dynamic base-age invariant site index models with the guide curve method (static models). For the evaluation of the candidate models qualitative and quantitative criteria were used. We also verified the stability of the candidate models, preferring a model providing fewer site class changes when predicting site index from different ages. The Lundqvist-Korf function fitted with the guide curve method proved to be effective and accurate for site classification and dominant height predictions of African mahogany stands. The range of observed site index, at a reference age of 15, was between 17 and 33 meters.