Regimes de manejo para Hovenia dulcis Thunb. em Floresta Ombrófila Mista como alternativa de controle e geração de rendas

Abstract

The exotic invasive specie Hovenia dulcis, is a contaminant of different vegetation types in southern Brazil such as: the Mixed Ombrophylous Forest (MOF), Semideciduous Forest and Grassy-Woody Steppe. Due to the threats for the natural balance of the biological communities and for the genetic resources of the country, control mechanisms of Hovenia dulcis must be sought in order to establish strategies to mitigate the adverse impacts on ecosystems. In this work were developed Hovenia dulcis management proposals under natural regeneration in MOF fragments in the South Central region of Paraná State, aiming to control the specie and at the same time generate income to landowners. For this, were enumerated all individuals of the trees (100% inventory) present in the forest fragments of 16 different landowners. In total 904 trees above 10 cm DBH were listed and formed the set of trees to be managed. The trees were grouped into diameter classes in order to serve mainly for the quantification of the current stock (measurement year - 2012) and for determined the trees sample for the complete stem analysis. For the current stock was determined the optimum cutting patterns of the stems individually (sawmill volume), in order to maximize the usable length of the stem. The optimization process generated an average use of the stem of 96.36% and a total volume for sawmill of 105.63 m3, just under a third of the total volume of 357,3 m3. During the development of the management proposals dendrometric models (hypsometric, volumetric and stem taper) and biometric (global growth models and models for individual trees) were used as tool, adjusted by two different methods: observed age (known age) (d = f (a)) and according to a scale of relative age (d = f (ra)). The database for the dendrometric models adjustment included 80 trees cubed and the biometric models were adjusted with 40 trees sampled by stem analysis technique. The adjusted models had appropriate and consistent results when compared with those usually obtained for species growing in native forests. The biometric models adjusted in function of relative age generated superior results with R 2 adj of 0.82 and Sxy% of 14.5 when compared to the adjustment by the know age method, which presented R 2 adj of 0.59 and Syx% of 22.1. Management proposals were based upon three different methods that aim to maximize revenue, with the selection of trees according to the projection of individual economic value over the planning horizon. The methods applied were: revenue maximization (MaxR), balanced revenue maximization (MaxR_B) and controlled revenue maximization (MaxR_C). For each of the three methods were simulated three different cutting cycle (2, 3 and 4 years), through the development and resolution of a specific linear programming algorithm per method. Because of the difference between the objectives of each method, an evaluation was performed for the three different cutting cycles. It was used technical and environmental criteria in assessing the best cutting cycle for each method. The cutting cycle every 2 years was superior for the technical criteria in two methods (MaxR and MaxR_B). On the other hand, the cutting cycle every 4 years was superior in environmental criteria for all methods. Those management proposals generated much better results when compared to results of a total harvest aimed the specie eradication in the year zero. The management options with the highest revenue and the largest total volume when compared with the result of the current stock was higher by 44 and 60%, for revenue and total volume, respectively. At the end, all management proposals fulfilled the main objectives initially set for this research.