Eficiência da determinação botânica de árvores no manejo florestal no estado do Pará.

Abstract

One of the main recognized aspects to achieve sustainability in forest management, biodiversity conservation, and the responsible use of natural resources in the Amazon is the improvement of botanical identification in forest inventories. In this context, the present work aims to evaluate the efficiency of tree botanical determination using dendrological methods and tools, contributing to the sustainable use of managed forests. The study area is Forest Management Unit (UMF) IV, located in the Paru State Forest (FLOTA) in the municipality of Monte Alegre. Botanical material collections were carried out for the common names listed in the Authorization for Forest Exploration (AUTEF) of Annual Production Unit (UPA) V. For each name, botanical samples were collected from at least five trees per species, and for the five names with the highest volume, collections were extended to the maximum number of trees harvested during the field team's period. The botanical samples had their scientific names determined based on morphological characterization and subsequent comparison with samples deposited in the Felisberto Camargo Herbarium of UFRA. With the obtained scientific names, the accuracy level of species determination was evaluated, grouped under each common name, and morphological differences of each species were compared to verify the agreement of the scientific names related by the company using the Kappa Coefficient at the family, genus, and species level. Of the twelve common names evaluated, five showed identification errors, with more than one species grouped under each common name. For example, under the names Ipês (ipê-amarelo and ipê-roxo), five species of Handroanthus were associated with the two common names, and under the name Currupixá, six species of Micropholis were recognized. It was observed that the errors in determination occur mainly at the species level, represented by the Kappa coefficient, which found moderate reliability (k = 0.413, agreement = 41%), while at the genus and family levels, the Kappa coefficient was extremely high (k = 0.985, agreement = 99%). Therefore, it was observed that the common names associated with trees during forest inventories group different species, which are not evidenced in the scientific name lists in management plans, reinforcing the need for improvements in botanical determination in forest management based on the morphological characteristics of each species.