ARCHIVAL WEBSITE
You are viewing the archived website of Pond Dynamics / Aquaculture CRSP. When using this website, please understand that links may be broken and content may be out of date. You can view more information on the continuation of PD/A CRSP research archived at AquaFish Innovation Lab.
Central Database Management
PD/A CRSP Eighteenth Annual Administrative Report

Previous Section Table of Contents Next Section


Central Database Management

MOU No. RD009G

Staff

Oregon State University, Corvallis, Oregon
John Bolte US Principal Investigator
Doug Ernst Database Manager (Research Associate)

Networking

CRSP Database staff John Bolte and Doug Ernst are developing a collaborative effort with research scientist L.G. Obaldo from The Oceanic Institute in Waimanalo, Hawaii, concerning shrimp culture, empirical datasets, and the development of component and system models. They are also collaborating with Kevin Hopkins, an aquaculture professor at the University of Hawaii, on statistical work and the production of overview-level journal articles regarding the database.

Ernst continues to serve as an online aquaculture extension agent, and he fields aquaculture-related questions from the general public and high schools. He receives one to two information requests per week.

Publication

Ernst, D., 2000. AquaFarm©: Simulation and decision-support software for aquaculture facility design and management planning. Ph.D. dissertation, Oregon State University, Corvallis, Oregon, 400 pp.

Award

Doug Ernst was asked to serve on an editorial committee of the journal Aquacultural Engineering.

Report: PD/A CRSP Central Database Management and Development

Ninth Work Plan, Database Management 2 (9DM2)

Douglas H. Ernst and John P. Bolte
Department of Bioresource Engineering
Oregon State University
Corvallis, Oregon, USA

Introduction

The PD/A CRSP Central Database is a centralized storage and retrieval system for aquaculture research data and related information (Hopkins et al., 1987; Batterson et al., 1991; Ernst et al., 1997; Bolte et al., 1998; Ernst and Bolte, 1999; Ernst and Bolte, 2000). The Database currently contains datasets from CRSP-sponsored research, but it is open to other aquaculture research with compatible objectives and compliance with standardized methodology. Access to the Database is available cost-free and is of interest to researchers, educators and students, outreach and extension agents, and producers in pond-based aquaculture. Datasets may be searched and retrieved based on specified location (research site), calendar year, fish species, fish culture methods, and desired type of data. An interface to the data and related information contained in the Database is provided at the Database website, located at <biosys.bre.orst.edu/crspDB/>. For intensive users of the Database, the contents are also available on electronic media (CD or Zip Disk).

The Database currently contains 108 aquaculture production studies and represents the world's largest inventory of standardized aquaculture data. The majority of studies in the Database are for production of Nile tilapia (Oreochromis niloticus) in sub-tropical and tropical solar algae ponds receiving inputs of plant materials, inorganic/organic fertilizers, and/or prepared feeds. Studies of other pond fishes and penaeid shrimp, under monoculture and polyculture management, are also available. Countries with research and research-support projects that have contributed to the Database include Egypt, Honduras, Indonesia, Kenya, Panama, Peru, Philippines, Rwanda, Thailand, and the US.

The Database project was started by the CRSP in 1985. The original purpose of the Database was oriented towards aquaculture science and engineering. Objectives were to:

1) Provide a mechanism for analysis of variance and multivariate analyses among geographically dispersed aquaculture research sites (in addition to analyses within single ponds and among ponds at a single location); and
2) Support development of predictive models for aquaculture pond processes and software for aquaculture design and management.

Since the placement of the Database on the Internet in January 1997, the purpose and functionality of the Database have expanded to address needs of aquaculture educators, planners, and producers. While continuing to serve as a centralized repository for aquaculture research data, the Database website also provides data search and extraction procedures, data graphing functionality, and contextual information specific to extracted datasets. (Additional statistical and regression tools for extracted datasets are under development.) Users can retrieve results of fish culture studies for a variety of geographical regions and production strategies, along with related information such as analytical methods, research site descriptions, publications, and author contact information. In this capacity, the Database serves as a major outreach mechanism of the PD/A CRSP.

Work Plan Tasks and Accomplishments

Objectives of the work to be completed under the Ninth Work Plan address the needs of both data suppliers and users. For data suppliers, objectives are to support the ongoing entry of new datasets into the Database, including continued improvements in administrative and technical support. For data users, objectives include continued improvements in Database query procedures, linkage support to related information resources for extracted datasets, statistical processing and data reduction, and provision of automated regression procedures to calibrate fish growth models that can be used as predictive tools for aquaculture planning and management.

Objectives of this report are to describe accomplishments to date and ongoing work for each of the tasks to be completed under the Ninth Work Plan for the reporting period 1 August 1999 to 31 July 2000. Database architecture, programming, and Internet software tools used in this work are described to a large degree in Ernst et al. (1997).

Database Hardware and Software

The hardware and software of the Database website server were upgraded to improve Internet security and website performance. Hardware upgrades included expanded storage, faster processing, and a write drive for compact disks. The latter provides a mechanism to provide large datasets to intensive users of the Database, as a supplement to Internet access. Current copies of the Database are also maintained on two additional computers, one of which has an automated daily backup system. Software upgrades included a new operating system (Windows 2000®; Microsoft©) and the latest versions of ColdFusion® (version 4.5; Allaire©) and Access® (version 2000; Microsoft©). In the course of this upgrade, the existing protocol used to communicate with the Database (ODBC) was changed to OLEDB, as recommended by Microsoft©. Unfortunately, the use of OLEDB caused unanticipated problems for both data retrieval and entry at the Database website due to high-level program bugs in the software products of Allaire© and/or Microsoft©. At the time of this report, some of these problems remain unresolved. If solutions cannot be found soon, alternative programming strategies will be required.

Independent of the OLEDB problem, alternative user interface programming mechanisms for the Database website were explored. While these alternative mechanisms are generally transparent to the user, the graphical interface options, data processing speed, and complexity of website programming vary with the methods used. Ongoing objectives are to provide the best possible performance for users while minimizing the complexity of website programming. Work accomplished in this area consisted of the development and assessment of a "client-based" approach (using Java®; Sun Microsystems©), as opposed to the "server-based" approach currently employed (Cold Fusion®; Allaire©). Given the considerably different program architectures required by these two approaches, staff compared the two approaches to determine which was most appropriate for the Database website. Results of these exercises showed that the client-based approach was excessively complex and required software products that were not sufficiently developed. The server-based approach in use was determined to be the best available technology. We will continue to periodically assess the methods used at the Database website relative to new developments in Internet technology.

Data Submission Templates—Economic Data

The data templates and specific data types used in the Database are described in the Data Submission Manual, available at the Database website. An updated version of this manual was published in June 2000, including changes to the Project Specification Form, added data types, and added data templates. The data templates and types of the Database have been relatively stable for some time, but continued progress was made in the area of economics. A data template was added to the Database for "full enterprise budgets," which complements the simpler templates for "partial enterprise budgets" added in the prior year. Economic data would be useful to supplement the wealth of physical, chemical, biological, and management data already available.

Partial enterprise budgets support comparative economic analyses of production trials (experimental treatments) in comparison to a base production scenario (control treatment). Partial budgets do not require comprehensive itemization and accounting of all cost and revenue items, given that their purpose is to compare the relative economics of alternative production practices. For example, in a pond production study comparing fertilization only to fertilization plus supplemental feeding, partial budgets could be limited to applied materials, fish produced, and labor, given that all other cost items are comparable. Partial budgets may be performed for on-farm production trials as well as research experiments. Given their simplicity, the Database Manager asks that all aquaculture production studies submitted to the Database include at least partial budget data.

The new data template for full enterprise budgets consists of:

1) Itemizations of the resources consumed and produced, including quantities and unit costs;
2) Additional cost items such as maintenance, depreciation, and interest; and
3) Derived values such as total costs, gross receipts, net gain/loss, and break-even prices and yields (kg ha-1 yr-1) for produced fish.

All quantities and costs are unitized to a per area (hectare) basis. Any number of cost and revenue items may be lumped together under "miscellaneous" cost and revenue quantities, respectively, at the discretion of the research project submitting the data. For full enterprise budgets to be most useful, however, they must contain all significant cost and revenue items itemized to a meaningful degree. The Database template used for full budgets supports complete flexibility in itemization, regarding both the extent of itemization and the specific items listed. Similar to the contextual information that is available for physical, chemical, and biological datasets from fish production trials, economic datasets are fully supported with contextual information such as farm description, production area, and technology used.

Data Submission Progress

Data submissions to the Database from CRSP research projects can be broken into two major periods: 1) First through Seventh and Interim Work Plans, or through 31 July 1996, and 2) Eighth and Ninth Work Plans, or from 1 August 1996. For the earlier period, data submissions for CRSP projects were not contractually required. Efforts to collect past-due data submissions for this period are ongoing and ultimately depend on the goodwill of CRSP researchers. From the Eighth Work Plan onward, data submissions from CRSP projects are contractual requirements.

Past-due datasets are welcome from the First through Fourth Work Plans (1982 to 1989) but are not being actively pursued. It is felt that too much time has elapsed to expect dataset submissions from this period. It is not known how many past-due datasets from this period would be appropriate for use in the Database.

An inventory was completed of the CRSP studies performed from 1989 to 1996 (Fifth through Interim Work Plans) and the associated datasets potentially due to the Database (Table 1).

Results showed that 39 out of 87 datasets have been submitted for this period. The remaining 48 of 87 (55%) studies not submitted represent a worst-case scenario, since of these datasets it is not yet known which are appropriate for the Database. Most of these past-due studies are somewhat ancillary to the major research thrust of the CRSP for this period. Efforts to determine the usefulness and availability of these datasets are currently underway and consist of correspondence with the specific project leaders who carried out the research. A final, concentrated effort to gain these past-due datasets will continue through the end of 2000.

Dataset submissions associated with the Eighth and Ninth Work Plans (1996 to present) are current relative to project schedules, orwhere notactive correspondence with project leaders is in progress. Correspondence among the Database Manager, the CRSP Program Management Office, and project leaders has documented exceptions that exist and intended submission schedules. We are pleased to report that we have received our first economic datasets for entry into the Database.

Work Plan and Project Summary Tables

As used here, the term "work plan" conforms to that typically used by the CRSP and the term "project" collectively refers to the individual studies and experiments performed within these work plans. The purpose and content of the work plan and project summary tables maintained in the Database were described in an earlier report (Ernst and Bolte, 2000). These tables contain specifications and related information for experiments and studies completed over the lifetime of the CRSP (1982 to present). In summary, they facilitate administrative and communication tasks of the Database Manager and provide contextual information for extracted datasets.

Beginning with the Eighth Work Plan, experiment-treatment specifications have been submitted by researchers in conjunction with project datasets (see Data Submission Manual). This has been working very well and has supported the development of both textual and codified experiment-treatment descriptions. Textual descriptions provide short narratives for review by Database users, and codified descriptions support data searches based on experiment methods and materials. Experiment-treatment specifications list and describe the experiment-treatment protocols used in fish production studies. This information essentially consists of the fish culture methods employed and is required to take full advantage of data search and extraction procedures at the Database website.

Prior to the Eighth Work Plan, materials and methods information was not provided with submitted datasets. To the extent possible, experiment-treatment specifications for past projects have been developed by reviewing project reports and compiling management data in the Database. Results, questions, and incomplete information identified by this work were reviewed in a prior Database report (Ernst and Bolte, 1999). To verify and augment this information, forms have been made available to researchers at the Database website for adding, editing, and reviewing the experiment-treatment specifications used in individual projects. CRSP researchers are being asked to access these forms for projects under their responsibility and make additions and corrections as necessary. This effort is ongoing and progress has been slow.

Dataset Searching, Processing, and Linkage

Database search and linkage procedures available at the Database web site feature 1) completely user-directed search criteria and 2) facilitated, context-specific linkage to related resources. These website facilities have been described in previous reports (e.g., Ernst and Bolte, 1999; Ernst and Bolte, 2000). Data search criteria include fish production site, dates, production methods and species, and desired data types. As discussed earlier, use of search criteria based on fish production methods is only as good as the available experiment-treatment specifications. While significant progress for past work plans has been made in gaining these specifications, continued effort is required.

Continued progress has been made in the ability to direct Database users to context-specific, Internet, or locally accessible information that is relevant to extracted datasets. The basic structure and operation of this information network remains essentially the same as that described in earlier reports. Entry of new information is ongoing. Major areas of information include:

1) Research site and facility descriptions;
2) Analytical methods used in production experiments;
3) Explicit definitions of all available data types, with links to analytical methods where appropriate;
4) Experiment-treatment specifications, available as concise narratives;
5) References to electronic and printed publications; and
6) Author citations and contact information.

These navigational linkages consist of automated active links and manual reference links. The end objective is to make all links active. These linkages take advantage of project information and literature maintained at the PMO website as well as new components of the Database developed for this purpose. The availability of these related information resources enhances the ability of Database users to interpret and apply extracted datasets.

Progress has been made in the development of statistical and regression tools for extracted datasets, but these tasks have not been fully completed and implemented at the Database website. Completion of these tasks has been hampered by unanticipated technical challenges encountered in the combining and processing of raw, replicate data into treatment summary statistics and regression parameters. This area of work is currently the top priority of the Database Manager and considered vital to the continued development of the Database. While aquaculture scientists and engineers may prefer raw data, they are of limited use to those working in aquaculture education, extension, and production and to collaborators operating related aquaculture databases. These users will be best supported by summary statistics and design and management tools generated from extracted datasets.

To address these needs, four tasks remain in progress. These consist of the development of automated procedures to generate:

1) Experiment-treatment summary statistics (range, mean, and variance statistics and support for analysis of variance);
2) Derived values useful to fish culturists (fish growth and feeding rates, feed conversion efficiency, and biomass density and productivity);
3) Calibrated models for describing and predicting fish feeding and growth rates based on specific datasets; and
4) Summary tables of the specifications and results of multiple treatments from multiple experiments.

Areas of these tasks that have been completed include:

1) Mechanisms to define and group experimental treatments;
2) Mechanisms to group replicates (ponds) into their original experimental treatments;
3) Mechanisms to coordinate time-series replicate data within a given treatment;
4) Development of models for fish growth based on fish size, water temperature, and availability of prepared (exogenous) and natural (endogenous) foods; and
5) Development of models for estimating natural foods based on primary productivity, whole pond respiration rates, and "critical standing crop" and "carrying capacity" fish biomass densities.

Database Use and Promotion

Since its inception in January 1997, the Database website has received a total of 4,680 visitors as of 31 July 2000. Over 1,700 of these visits occurred in the last year, and the usage rate of the Database website continues to accelerate. Improved tracking of Database users has been implemented, consisting of hit counters distributed throughout the Database website to monitor the relative use of the seven major areas of the site. Methods to automatically log country of origin and other general characteristics of website visitors are under development. The use of optional user profile forms (e.g., type of aquaculture work and specific interests) is under consideration. Objectives are to better understand user needs.

Promotion of the Database to potential users is ongoing. Critical issues are awareness of the Database availability, content, and applications. To date, the Database has been promoted through aquaculture conferences, publications, and linkages to related websites and databases. Direct promotion to specific user groups remains to be accomplished, following completion of the statistical data summary and reduction tools discussed earlier and a supporting Database Users Manual. Data submission from outside the CRSP has been limited, but the potential for this type of collaboration will increase as the analytical capacity available at the Database website is further developed and made known.

Collaboration with related databases has been established for some time. Collaborators include the International Center for Living Aquatic Resource Management (ICLARM), Network of Aquaculture Centres in Asia-Pacific (NACA), and Consortium of International Earth Science Information Networks (CIESIN). Data content and format requirements have been established for these potential indexing links and access points to the Database, but completion of the intended scope of these collaborations will follow completion of the statistical data summary and reduction tools discussed earlier. Simple site-to-site web linkages with major fishery, aquaculture, and agriculture professional societies, research groups, universities, and information providers continue to be developed.

Literature Cited

Batterson, T., H. Berkman, K. Hopkins, R. Piedrahita, and T. Popma, 1991. Final Report on Database Management. Pond Dynamics/Aquaculture CRSP, Oregon State University, Corvallis, Oregon, 51 pp.

Bolte, J.P., D.H. Ernst, and D. Lowes, 1998. Central Database Management. Fifteenth Annual Administrative Report. Pond Dynamics/Aquaculture CRSP, Oregon State University, Corvallis, Oregon, pp. 36–45.

Ernst, D.H. and J.P. Bolte, 1999. PD/A CRSP Central Database Management and Development. Sixteenth Annual Administrative Report. Pond Dynamics/Aquaculture CRSP, Oregon State University, Corvallis, Oregon, pp. 9–14.

Ernst, D.H. and J.P. Bolte, 2000. PD/A CRSP Central Database Management and Development. Seventeenth Annual Administrative Report. Pond Dynamics/Aquaculture CRSP, Oregon State University, Corvallis, Oregon, pp. 11–15.

Ernst, D.H., J.P. Bolte, and D. Lowes, 1997. PD/A CRSP Central Database: An information resource for pond-based aquaculture. In: K. Fitzsimmons (Editor), Tilapia Aquaculture, Proceedings from the Fourth International Symposium on Tilapia in Aquaculture (ISTA), November 1997. NRAES-106, Vol. 2. Northeast Regional Agricultural Engineering Service, Ithaca, New York, pp. 683–700.

Hopkins, K.D., J.E. Lannan, and J.R. Bowman, 1987. A database management system for research in pond dynamics. CRSP Research Report 87-1. Pond Dynamics/Aquaculture CRSP, Oregon State University, Corvallis, Oregon, 4 pp.

Piedrahita, R.H., C. Boyd, and J. Szyper, 1991. Handbook of Analytical Methods. Pond Dynamics/Aquaculture CRSP, Oregon State University, Corvallis, Oregon, 150 pp.

Previous Section Table of Contents Next Section