Agronomy Journal 94:1163-1171 (2002)
© 2002 American Society of Agronomy
SOYBEAN
Deltasoy—An Internet-Based Soybean Database for Official Variety Trials
Lingxiao Zhang*,a,
Wanwen Qib,
Ling Sub and
Frank Whislerb
a Delta Res. and Ext. Cent. and Dep. of Plant and Soil Sci., Mississippi State Univ., Stoneville, MS 38776
b Dep. of Plant and Soil Sci., Mississippi State Univ., Stoneville, MS 38776
* Corresponding author (lzhang{at}drec.msstate.edu)
Received for publication June 19, 2001.
 |
ABSTRACT
|
|---|
Official variety trial (OVT) information in soybean [Glycine max (L.) Merr.]-producing states is critical to soybean growers in decision-making processes. Most soybean-producing states in the USA have their own soybean research web sites with information relating to the soybean variety trials. A web-based database system has been developed that can help soybean growers and researchers access the most recent OVT information quickly and in response to their specific interests. This new system also summarizes past research results systematically and has the potential to link ongoing research in related areas of soybean production, modeling, and other research efforts. Microsoft Access, along with other computer languages [C++, common gateway interface, Standard Query Language (SQL)], were used in building this database. The database uses Mississippi OVT data, including yield, location, and disease information. This database has the potential to extend its capability by adding more information from variety trials of other states and other crops. Therefore, it should be a helpful tool for soybean producers and researchers in Mississippi as well as other interested parties.
Abbreviations: DBMS, database management system • MG, maturity group • OVT, official variety trial • RR, Roundup Ready
|
INTRODUCTION
|
|---|
SOYBEAN official variety trials (OVTs) have been conducted in Mississippi for more than 20 yr. As in most other soybean-producing states in the USA, OVTs are one of the most important information resources for soybean producers, seed suppliers, and researchers in their decision-making processes.
Most soybean-producing states have published OVT results on the Internet. However, information from most OVTs presented on the Internet is either presented flat (documents and tables) or in some other noninteractive format. While those formats may be clear to read with a general purpose, they do not have the needed flexibility when more specific information is needed. Sometimes the documents and tables become so large that it is difficult to find the specific information sought.
A database system provides universal control of its operational data (Date, 1982). Compared with nondatabase systems, database systems have several advantages, including simplicity, specificity, flexibility, and speed. They provide the users with the targeted information through numerous sets of data in a large database. They have been applied in almost every area, including agriculture (Van Evert et al., 1999a, 1999b).
In the last 10 yr, computer technology and information systems, and especially the World Wide Web system and the Internet, have affected many aspects of people's lives. Because of the advantages of the web and database systems, it could be very useful to combine these two technologies into one and build up a web-based database for OVT data. The web-based database system can provide a facility for storing data more safely, accessing data more readily, and sharing data more promptly. Deltasoy, an Internet-accessible database system for the state soybean OVT data, was designed and built based on this philosophy. The objective was to provide soybean growers and researchers with quick and easy access to OVT information. Deltasoy can also record and summarize OVT results in a systematic way so that data can be traced back when historical information is needed. With this database, research scientists may use the OVT data dynamically to identify potential new problems or research directions. Furthermore, because of its database format, the OVT results may also be linked with other ongoing crop research areas, such as diseases, phenology, modeling, and Global Positioning System (GPS) or Geographic Information System (GIS) work. If this database can be used successfully with soybean variety trial data, it could also serve as a model for providing OVT information for other crops.
|
DESIGN AND PROCEDURES OF DELTASOY
|
|---|
Deltasoy, a web-based database system, was developed to help soybean growers and researchers access the most recent information on soybean production and state variety trials. The main function of this database is to search certain parameters (or factors) based on the user's specific queries. Specific parameters (or query items) can be submitted to the database, and the database system will provide feedback to the user's request. The contents of the system include two major components: agronomic (independent and dependent variables) and computer and information systems (web interface, computer programming, data communication, etc.).
Determination of Agronomic Parameters and Their Relationship
The primary data sources for this database are from the Mississippi OVT reports (Askew et al., 1997a, 1997b; White et al., 1999, 2000, 2001) and Arkansas Soybean Performance Test (Dombek et al., 1999). Determination of parameters in this database is based on current flat-form tables of the Mississippi OVT reports plus factors that may be involved in future extensions of this database.
Agronomic Parameters
Some of these parameters are categorical while others are numerical:
Year. In the database, a user can search for the results for one particular year or from a range of multiyear combinations by selecting the range of years and an average function. Users may also look at the variability (or stability) over several years. Currently, there have been 6 yr (1996–2001) of data from soybean OVT from Mississippi and 1 yr (1999) of data from Arkansas converted into the database system.
State. All soybean-producing states in the USA have been listed. However, at this time, only data from Mississippi and Arkansas have been included. Efforts are underway to extend this database into a multistate functional database project.
Maturity groups. Theoretically, soybean maturity groups (MGs) range from 000 to X. In reality, soybean breeders usually use a more accurate system (with one decimal point) to describe soybean maturity (from 0.1–10.9). In this database, MG 000 is set between 0.1 (early MG 000) and 0.3 (late MG 000), MG 00 is set between 0.4 and 0.6, and MG 0 is set between 0.7 and 0.9. Then, MG I is set between 1.0 and 1.9, and subsequent MGs are set using the same principle. However, in each state, the MG range adapted is different. Therefore, users can select their range interest. For example, the MG range of Mississippi and nearby states is most likely to be from III to VI. In recent Mississippi OVT tests (up to 2000), MGs used ranged from 3.4 to 6.9.
Variety. Because new varieties are released so rapidly, it is hard to list all of the varieties using a selection function. Therefore, this category has been designed as a type-in function. However, efforts have been made to provide the users with a variety table, arranged either by company entry or by alphabetical list. Currently, the table only includes varieties that have been used in the Mississippi and Arkansas OVTs.
Roundup Ready vs. conventional. The number of Roundup Ready (RR) soybean varieties has steadily increased for the last 5 yr (Zhang et al., 2001). In Mississippi OVT, the number has increased in the last 5 yr from less than 5 to 80% of total number of commercial varieties available. In this database, the RR soybean varieties are separated from the conventional ones. Most other soybean-producing states also conduct their OVTs by separating RR soybean from conventional soybean.
Irrigation. The experiments conducted in OVT tests contain two broad categories, irrigated and nonirrigated. Yields from most variety trials under these two conditions are greatly different when soil water is limiting. Because the main function of the database is to show the highest-yielding varieties, it is appropriate to divide the irrigated and nonirrigated experiments into two different categories.
Location. Individual locations are listed with the selection format. However, they will be different for each individual state. In the Mississippi OVT, the locations can be divided into two distinguishable production regions: Hill and Delta. Experiments on the Hill sites are usually all nonirrigated while most of the Delta locations are irrigated. Locations may also vary from year to year. In the current database, the latitude and longitude of sites are not recorded, but they may be added when the database includes more states.
Disease resistance information. Common soybean diseases (including nematode) are listed as a select function. Users may choose to eliminate a specific disease from the search results. Disease resistance and soybean nematode information have been obtained from experiments distinct from OVT tests and are incorporated into this database as a special function (further discussed in the section "Determination of Relationships between Variables").
Other factors. Soil types, plant height, lodging score, and maturity data are also listed. Other factors, such as shattering score, may be added to the database later.
Determination of Relationships between Variables
In this database, 13 major variables (attributes or entities) are included. These variables are divided into two categories: genotypic factors (MG, disease information, and RR) and environmental factors, which subdivide into phenotypic factors, or outcome (yield, plant height, and maturity date) and nonbiotic factors (year, location, irrigation, state, and soil type). One additional variable, variety, is used as the primary key (the attribute that is uniquely identified by the rows in a main table) to connect all of these variables. An entity relationship diagram (Fig. 1)
, based on the above category system, was used to design the Deltasoy database. The relationships among these parameters are normalized (data redundancy was reduced) in a dependency diagram (Fig. 2)
. Except for the primary key and foreign key (the same attribute as primary key, except used to link into another table), the two sets of attributes are not related. Consequently, each search with a MG or RR condition will require a join, a connection between two tables, which is quite common in practice. The relationship of disease and yield has been renamed in a changed dependency diagram (Fig. 3) . If a user is interested in the disease resistance information of certain varieties, up to three diseases can be selected in query boxes along with other queried parameters. After submitting the query and getting the results, a disease search program can be called by clicking on the variety name in the result set. The search selects the corresponding variety in the disease table. Disease information is not required and will not show up unless the user selects this function. Also, when all three boxes are selected with different diseases, it is likely that not many varieties will be shown in the resulting table.
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 3. Dependency diagram of parameters in Deltasoy database with changed relations between disease and yield.
|
|
Hardware and Software of Deltasoy
A web database is a collection of data that is accessed via a query language or by programming an application programming interface (Hobuss, 1998). A web-based database includes two components: web and database. The web component includes two distinguishing parts: hardware and software. The hardware consists of a server machine, network facility, power, and a database server. The software includes the system architecture, operating system and web server, interface design, database management system (DBMS), and development tools. Hardware and software can communicate with each other through a common gateway interface. Hardware, software, and common gateway interface are connected in a systematic way and form a single unit to function as an Internet database server (Fig. 4)
.
Hardware
Web server. A personal computer with a Pentium II system usually is capable enough to host a web database, especially when it is not expected to have a large number of accesses (say more than 20) in the same time period. Otherwise, a high-capacity work station should be considered. In this project, a Dell1 personal computer with 333 MHz Intel Pentium II processor is used as the Web server mainly due to its ease of maintenance.
Database server. The database server is used to run the DBMS. An economical way to build a web database is to use the web server as the database server. The advantage of sharing the web server with the database server is that the cost of hardware and communication is lower. The disadvantage is a decrease in speed. If high-volume access is expected, a set of multiple central processing units (CPUs) or clustered work stations with high-speed connections should be used. We did not expected to have many accesses at one time; therefore, the web server was also the database server.
Backup system. A database system needs a backup device for keeping files safe and data consistent and avoiding loss of data in case of natural disasters and human errors. The most common backup devices are hard drives and tapes. In this database, the data will not be changed when the web site is online, so a very simple backup is used, implemented by a backup macro embedded with Visual Basic scripts to copy the Deltasoy database system to deltasoy.bak. Only the database and the files related to the web site are backed up onto another machine. In case a failure occurs, these can then be copied to a new machine.
Other network facilities. The web server must be connected to the Internet, and each web server must have an Internet Protocol (IP) address, which is expressed in a numerical format. However, users usually use the web address, which uses alphabetic letters, to visit the sites. The Deltasoy web address is http://www.deltasoy.msstate.edu (verified 5 June 2002).
Software and Web Interface Design
Operating system and server software. The choice of operating system depends on the hardware being used. If a web server machine is built with a personal computer, Windows (95 or above) or Linux may be chosen. Our server uses the Windows NT system.
Database management system. For this database, Microsoft Access was selected as the DBMS. Microsoft Access is easy to use, the volume of data for storage is relatively small compared with some commercial databases, and many simultaneous accesses are not anticipated. The data is updated each year. A typical soybean variety trial may contain an average of 250 or fewer entries (including all MGs, conventional or RR) over 10 locations (including different soil types and irrigation status in each year). Tracking the next 20 yr of variety trial data from 30 potential soybean production states, the potential tuples, the data entry, would be about 1.5 x 106. The actual number of tuples involved should be much less than this number. Microsoft Access provides a very user-friendly interface and can also handle this estimated number of tuples.
Administration Tools and Procedures
The Windows operating system can be divided into two classes according to functionality: server (host side) and client (user side). On the web page, the back-end database is read-only to users. That means the Internet user cannot modify the content of the database. However, administration tools are needed for authorized persons to use and maintain the database.
|
AN APPLICATION EXAMPLE OF DELTASOY
|
|---|
A well-designed database can help growers make good decisions on variety selection, production practices, and marketing. It also can be a useful tool for soybean scientists to conduct research and extension work. The following example shows how this database can be used simply and effectively in a real-world situation. The circumstances in this example are hypothetical, but they are based on a realistic situation.
Problem Statement
A farmer near Stoneville (location), MS (state), needs to make some decisions for his soybean production plan in 2001. He has 200 ha each for both irrigated and nonirrigated (irrigation condition) fields. His fields are heavy clay soils (soil type). From past experience, he prefers to use RR (soybean type) soybean varieties. Before this time, he had only grown soybean varieties with MG V. He would like to know the potential of earlier soybean MGs because he has heard some information about earlier MGs from his neighbors. He decides to use Deltasoy to study some possibilities. He has the following three questions in mind:
Questions or Objectives
- Should he continue to plant all of his fields with the same variety and MG (V)?
- If not, which MG should he choose from (between MG IV and V)?
- Should he try some soybean varieties with earlier maturity than MG IV?
Procedures
First, the farmer (referred as the user in the following discussions) would open the Deltasoy homepage and go to Search. After clicking on Search, a U.S. map with each state clearly separated would be shown. An alphabetical list of the soybean production states would also be provided so that the user may choose his intended state either from the map or from the list. By further clicking state, Mississippi, an interactive interface would be provided.
Second, the user would input his situations (query conditions) into the query boxes. Assuming the user would like to look at results from the previous year (2000) first, he would select Year 2000. Otherwise, the user could select Average on the first box and then select the second box with a range, assuming a 3-yr result (from 1998–2000). The completed inputs would be listed as in Table 1.
After inputting all of these query conditions, the form would be shown as in Fig. 5
. Then the user could click on Search at the bottom of the page, and it would provide an output result shown as in Fig. 6
. If the user would like to look at the nonirrigated condition, he would reclick at the Yes box of the irrigation attribute to take the check mark off and click the No box again to select the nonirrigated condition. Then the user could click on Search again. A new result would be generated shown as in Fig. 7
. The user then can review the current results and compare them with the previous results.
If the user was interested in looking at the potential of even earlier-maturity soybean, then the only thing the user would need to do would be to change the query condition from 3.0 (instead of 4.0) to 6.0 at the MG attribute. The results would be the same as before under irrigated condition (data not shown) but quite different under nonirrigated condition (Fig. 8)
.
|
RESULTS AND DISCUSSION
|
|---|
The mid-South climate was very dry during the 2000 growing season, and yields from irrigated and nonirrigated conditions differed greatly (Fig. 7 and 8). Furthermore, among those top 10–yielding cultivars, nine had MG values between 5.0 and 5.9. Only one cultivar was between 4.0 and 4.9. On the other hand, the situation was reversed under nonirrigated condition. The number of cultivars with an MG value between 4.0 and 4.9 was 9 out of 10. The one cultivar not in this range was close to the high end of MG IV (at 5.1). So, to answer to the first question, the farmer should NOT grow soybean on all of the irrigated and nonirrigated fields with the same cultivar. To answer the second question, he should use MG IV (4.0–4.9) cultivars on his nonirrigated field but keep MG V soybean on his irrigated fields because irrigated MG V (5.0–5.9) cultivars showed greater potential for a high yield.
When query condition at lower end of the MG value changes from 4.0 to 3.0 (query MG value using 3.0–6.0), six cultivars with MG value between 3.0 and 3.9 were in top-10 listing under nonirrigated fields (Fig. 8). The rest are in the range 4.0 to 4.9. This indicates that early-maturity soybean might perform well in nonirrigated fields under these circumstances though some other conditions such as planting dates might be a little different. Therefore, to answer the third question for this farmer, he should not grow earlier MG soybean on his irrigated field, but he should probably try some earlier soybean on his nonirrigated field.
Of course, making cultivar selection and planting decisions is always difficult for soybean farmers. Depending on a one-dimensional resource, such as Deltasoy, to make an important decision is very risky. Other factors may also be involved. Therefore, consulting soybean specialists from public sectors and private companies is important and highly recommended. Information systems such as Deltasoy are for decision support, not for actually making the decision.
|
POTENTIAL AND SIGNIFICANCE OF DELTASOY
|
|---|
The above example illustrates a simple production decision by a single user. Of course, 1-yr data should not be the sole basis for a critical decision, but the principle of the application is the same. The user could switch to a different year to look at different weather or examine multiyear averages to consider the stability and variability across several years. By changing query conditions, the user can get specific results quickly.
New information from soybean research is aiding the soybean grower in making better decisions. Compared with a flat table format, this database system has the advantages of simplicity, user friendliness, speed, flexibility, and extensibility. Deltasoy has been in use since 1999. Currently, 5 yr of data from Mississippi (1996–2000) and 1 yr from Arkansas (1999) have been entered. In 2000, Mississippi OVT data were accessible from Deltasoy in late October, almost 2 mo earlier than most other states. Positive responses have been received from those who have used Deltasoy. This project is useful to Mississippi soybean farmers and researchers and also can have tremendous impact on how people look at the application of new information to science and research.
|
FUTURE WORK NEEDED
|
|---|
Deltasoy is an Internet-based database system. The database and web technology involved in this project surpass the old methods in many aspects. Deltasoy provides easy and fast access to the latest data from soybean cultivar trials. The preparation and distribution of the soybean cultivar trial bulletin required months. Via the web database, the data is available to farmers and researchers within a month. Through the powerful search form, a user can dynamically query soybean cultivar trial data as well as statistical information.
Because this database application is a relatively new approach to utilize the state soybean OVT information, many areas are still open for exploring and improving. Thus, further research works are needed in many aspects. Currently, we are focusing on how we can (i) include more functions to help soybean producers answer more comprehensive questions about cultivars, yields, and production issues; (ii) use this database more efficiently as a research and extension educational tool; (iii) link it to other ongoing research projects, such as GPS or GIS works in soybean productions and soybean modeling because the state OVT information may become one of the important components in these researches; and (iv) extend this concept and idea and create more Internet-based databases to provide the information more efficiently for other crop cultivar trials, such as cotton (Gossypium hirsutum L.), rice (Oryza sativa L.), wheat (Triticum aestivum L.), turfgrass, etc. We believe that when more data from other states are compiled into this database, it will become a powerful tool to provide more dynamic information for growers, extension educators, and researchers.
|
ACKNOWLEDGMENTS
|
|---|
We sincerely thank the Mississippi Soybean Promotion Board for their financial support to this project. We also appreciate the technical contribution from Mrs. J. Lu and J. Zhang and manuscript review and comments from Dr. C. Watson and Dr. J.D. Hesketh.
|
NOTES
|
|---|
1 Authors do not endorse any brand or products by mentioning the name of a company in this paper. 
|
REFERENCES
|
|---|
- Askew, J., A. Blaine, F. Boykin, R. Dobbs, F. Hancock, D. Ingram, R. Martin, D. Reginelli, A. Smith, T.R. Vaughan, and M. Young. 1997a. Soybean variety trials 1996. Inf. Bull. 315. Mississippi Agric. and Forestry Exp. Stn., Mississippi State.
- Askew, J., A. Blaine, D. Ingram, R. Ivy, W. Marlow, D. Shaw, A. Smith, C. Watson, B. White, and M. Young. 1997b. Soybean variety trials 1997. Inf. Bull. 328. Mississippi Agric. and Forestry Exp. Stn., Mississippi State.
- Date, C.J. 1982. An introduction to database systems. Addison-Wesley Publ. Co., Reading, MA.
- Dombek, D.G., D.K. Ahrent, R.D. Bond, and I.L. Eledridge. 1999. Arkansas soybean performance test 1999. [Online]. Available at http://www.uark.edu/depts/agronomy/dombek/soybeanreport99.pdf (verified 2 May 2002).
- Hobuss, J.J. 1998. Building Oracle web sites. Prentice Hall PTR, Upper Saddle River, NJ.
- Van Evert, K. Frits, E.J. Spaans, S.D. Krieger, J.V. Carlis, and J.M. Baker. 1999a. A database for agroecological research data: I. Data model. Agron. J. 91:54–62.[Abstract/Free Full Text]
- Van Evert, K. Frits, E.J. Spaans, S.D. Krieger, J.V. Carlis, and J.M. Baker. 1999b. A database for agroecological research data: II. A relational implementation. Agron. J. 91:62–71.[Abstract/Free Full Text]
- White, B., A. Blaine, D. Ingram, R. Ivy, W. Marlow, A. Smith, C. Watson, M. Young, and L. Zhang. 1999. Soybean variety trials 1998. Inf. Bull. 346. Mississippi Agric. and Forestry Exp. Stn., Mississippi State.
- White, B., A. Blaine, R. Ivy, W. Maily, W. Marlow, A. Smith, C. Watson, M. Young, and L. Zhang. 2000. Soybean variety trials 1999. Inf. Bull. 363. Mississippi Agric. and Forestry Exp. Stn., Mississippi State.
- White, B., A. Blaine, R. Ivy, W. Maily, W. Marlow, A. Smith, C. Watson, M. Young, and L. Zhang. 2001. Soybean variety trials 2000. Inf. Bull. 374. Mississippi Agric. and Forestry Exp. Stn., Mississippi State.
- Zhang, L.X., D. Poston, and F. Cooke, Jr. 2001. Survey of recent soybean production trends in the Mississippi Delta. In 2001 annual meeting abstracts [CD-ROM]. ASA, CSSA, and SSSA, Madison, WI.
TOP
ABSTRACT
INTRODUCTION
