Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/20/1998
Publication Date: N/A
Citation: N/A Interpretive Summary: In a previous paper we described the development of a database for the systematic storage of data, and all relevant additional information, from field experiments. Successful development of such a database requires that it be tested, to determine its usefulness and to find and correct weaknesses. In this paper we describe implementation of the database and show examples of its use. We used it to store all field data from a large multi-year, multi-investigator field experiment conducted by the USDA-ARS and the University of Minnesota to study agricultural impacts on water quality. We also used structured query language (SQL) to retrieve specific detailed subsets of the data in order to address specific research questions. The database design was found to be well-suited to these tasks, and we conclude that it will be particularly valuable in experiments such as these, where large volumes of data must be stored for extended periods, and subsequently analyzed by multiple researchers.
Technical Abstract: In this first paper of this series, we developed a data model for agro- ecological research data. The objective of the work described in this paper was to determine the practical usefulness of a database implemented from that data model. We used a commercially available relational database management system, and, generally, mapped the entities of the logical data structure to tables, entity attributes to table columns, and entity relationships to foreign keys. However, the entity representing measurement values was mapped to two separate tables, both to reduce the size of the database and to reduce the response time of queries involving this entity. We found that loading data into the database was the most significant hurdle to its use and developed a set of stored procedures that function as a data input language. The input language and other methods were used to load the data from an intensively monitored, multi-site, multi-year experiment into the database. The database was used to manage, explore and analyze the data from these experiments, as well as to share the data between collaborators and with others in an effective way. We conclude that a database implemented from the previously designed data model is a practical and extremely useful tool for research.