The development of a forestry geodatabase for natural forest management plans in Costa Rica

Forest management data is available for many tropical countries, representing a large and spatially rich source of tree level data. Over the past decade, we have collected and digitized over 1000 Costa Rican natural forest management plans, spanning 30 years and spread out over approximately 26,700 km2 along the country's Atlantic coast, northern lowlands, and southern Pacific coast. In order to analyze this unique collection of tree inventory data we developed a system to ensure the standardization, quality control, and reliable management of the dataset. We developed a relational geodatabase, forming logical associations between and within the spatial and tabular components of the forest management data. In this study, we outline the data standardization procedures established to permit the comparison of data across time and space. Further, we describe quality checks built-into the database's functionality to identify and reduce the presence of errors. The final customized forest management geodatabase efficiently stores a large and diverse dataset through the integration of logical relations, quality checks, and flexible data access across 32 tables and over 250,000 tree records. Through the use of quality tests, the database provides a means to improve overall data integrity and illustrates the magnitude, sources, and types of errors present in the initial dataset. Finally, the value of a comprehensive database for the management of forest data is demonstrated through an exploratory analysis of standardized tree taxonomic information. With this analysis, we begin to explore the potential strengths, weaknesses, and applications of forest management data for future ecological studies (e.g. species diversity assessment). For instance, although most of the forest management data (tree inventories and censuses) is collected using common names in the field, our standardization process has allowed us to depict trends similar to those found in ecological studies (e.g. dominant species for different ecosystems). Overall, our forestry geodatabase represents the most complete record of natural forest management practices in Costa Rica to date.