![[Board accepted]](http://www.istl.org/graphics/b2.gif){kind=small}
Agriculture Journal Literature Indexed in Life Sciences Databases
Jodee L. Kawasaki
Associate Professor
Information Resources Development Team Leader
Montana State University - Bozeman
alijk@montana.edu
Abstract
A variety of life sciences databases are available for libraries to purchase and for users to search. The purpose of this study was to discover which of these databases cover the agricultural journal literature for comprehensive searching and results. I compared the seven life sciences databases against the primary agriculture journals. The comparison results were mixed as to the databases' subject coverage. CAB Abstracts, an agricultural information database, had the best results; however, other databases with specific agriculture subject coverage were interspersed among more general life sciences databases when the total results were compared. The complete list in order from the most comprehensive coverage to the least coverage is CAB Abstracts, Web of Science, AGRIS, BIOSIS, Agricola, CSA Life Sciences, and Biological & Agricultural Index.
Introduction & Literature Review
With the pace of the world doing anything but slowing down, users want to design comprehensive searches quickly. Librarians need to know which databases cover agricultural literature completely to provide the best assistance available to users. There are many life sciences databases covering some or most agricultural literature. The choices can be overwhelming for users if many are available. If few databases are available, knowing which one will give the best results is important. From a collection development viewpoint, knowing the database with the most comprehensive coverage of agriculture journals may provide cost savings in the future. With tight budgets now the norm, one must consider which database gives the best agriculture journal coverage for budget reduction scenarios.There is a small body of literature indicating which serials are relevant in the agriculture sub-disciplines such as Dote and Letnes' (1996) agricultural economics study and Kelsey and Diamond's (2004) forestry study. Another article focuses on the single subject of sustainable agriculture (Weintraub 1997). A previous study by the author (Kawasaki 2002) compared the agricultural indexes of Agricola, CAB Abstracts, Biological and Agricultural Index, and Biological Abstracts, which looked at how the indexing of core serials changed from 1998 to 2002. A study comparing serial titles in current awareness services (Murphy 1996) indicates that each current awareness service's serial list was compared against the other services' list. These two studies are similar in nature, yet focused on specific issues.
The purpose of this study was to compare life sciences databases coverage of primary agriculture journals.
Method
For the purpose of this study, the author used an established list of agriculture journals considered core from the Literature of the Agricultural Sciences series edited by Wallace Olsen, et al., and published by Cornell University Press between 1991-1996. The Literature of the Agricultural Sciences series covers the sub-disciplines of agriculture: agricultural economics, soil science, animal science, crop science, agricultural engineering, forestry, and food sciences. The next step was to update the list as of 2003 to include title changes, splits, mergers, and annuals. Titles that ceased prior to 2003 were excluded from the journal list. The book from Hutchinson and Greider (2002) was consulted to include titles in the primary list that had not been previously considered core in the Olsen series. The primary agriculture journals list was confirmed through searching OCLC's WorldCat, Ulrich's Periodicals Directory, and BIOSIS Serial Sources. The list yielded 542 agriculture journals used for the comparison.This study analyzes seven life sciences databases for agriculture journals titles both scanned and indexed. The seven life sciences databases were chosen because of the life sciences or agriculture focus: Agricola from the U. S. National Agricultural Library (NAL); Biological and Agricultural Index Plus from H.W. Wilson Co.; Biological Abstracts (Biosis) from Institute for Scientific Information (ISI); CAB Abstracts, from CAB International (CABI); Cambridge Scientific Abstracts (CSA) Life Sciences Collection from CSA; and Sciences Citation Index Expanded, also known as the Web of Science from ISI. Other databases were investigated for inclusion: PASCAL from Institut de l'Information Scientifique et Technique (INIST) of the French National Research Council (CNRS); Food Science and Technology Abstracts from International Food Information Services; and Zoological Record from ISI. The author excluded these three databases since each had a specialized emphasis that resulted in a small, focused total of agriculture journal titles scanned or indexed.
The lists of which journals are indexed for each of the seven life sciences databases were compared to the agriculture journals list. The criterion was the titles had to match. The author performed a confirmation check by searching the ISSN of the agriculture journals within the database. Titles may be duplicated, but ISSNs are unique. The extra measure of checking ISSNs provided accurate information for matches. When an ISSN was not found, publisher or publication place was compared for verification purposes. These steps improved the accuracy of the comparison study.
Results
Out of the 542 agriculture journals, only 24 titles (4.4%) were indexed elsewhere than these seven databases. Of these 24 titles, Food Science and Technology Abstracts and Zoological Record each indexed one unique title. The other 518 journals (95.6%) were either scanned or indexed by the seven life sciences databases. The complete comparisons are shown in Tables 1-3 and discussed below.Table 1 gives total number and percentage of primary agriculture journals found in each database. Some titles were not found as scanned or indexed by any of these life sciences databases, the numbers and percentage are also given in the table.
Table 1. Primary Agriculture Journals in Each Database
2003 | Total N=542 | Percentage of N | Unique | Percentage of N |
| Agricola | 259 | 47.6 | 0 | 0 |
| Agris | 340 | 62.7 | 0 | 0 |
| BAI | 116 | 21.3 | 0 | 0 |
| Biosis | 311 | 57.2 | 2 | 0.04 |
| CAB | 499 | 92.1 | 41 | 7.6 |
| CSA | 219 | 40.3 | 0 | 0 |
| WoS | 402
| 73.9 | 10 | 1.8 |
| Other | 24 | 4.4 | 24 | 4.4 |
The ranking of the seven life sciences databases with respect to how many primary agriculture journals are contained in each one is CAB Abstracts (CAB) covering 92.1% of the primary agriculture journals, Web of Science (WoS) indexes 73.9%, Agris includes 62.7%, Biosis covers 57.2%, and the rest are below 50% coverage of the primary agriculture journals. Agricola includes 47.6%, CSA Life Sciences (CSA) has 40.3%, and Biological & Agricultural Index Plus (BAI) covers 21.3% of the primary agriculture journals. There are 4.4% of the primary agriculture journals not indexed in any of these major life sciences databases. These 24 titles are specific to a different discipline such as medicine or economics, but apparently have some articles related to agriculture, enough so as to be listed as a primary agriculture journal. The author has no explanation as to why they are not indexed by one of the major life sciences databases.
Databases combined for searching | Total N=542 | Percentage of N | Missed Titles | Percentage of N |
| CAB and WoS | 514 | 94.9 | 28 | 5.1 |
| CAB and Biosis | 506 | 93.4 | 36 | 6.6 |
| CAB and Agris | 502 | 92.6 | 40 | 7.4 |
| CAB and CSA | 501 | 92.4 | 41 | 7.6 |
| WoS and Agris | 456 | 84.1 | 86 | 15.9 |
| WoS and Biosis | 430 | 79.3 | 112 | 20.7 |
| WoS and Agricola | 428 | 79.0 | 114 | 21.0 |
| WoS and CSA | 422 | 77.9 | 120 | 22.1 |
| Agris and Biosis | 404 | 74.5 | 138 | 25.5 |
| Agris and CSA | 376 | 69.4 | 166 | 30.6 |
| Agris and Agricola | 367 | 67.7 | 175 | 32.3 |
| Biosis and Agricola | 365 | 67.3 | 177 | 32.7 |
Table 2 indicates combined coverage of the primary agriculture journals. The most unique titles are indexed in CAB and WoS. Searching both of these databases would cover 514 primary agriculture journals, missing only 28 journals. Of these 28 missed, 24 are elsewhere, two are in Biosis, one is found in Agris and CSA, and one is found in Biosis and CSA. Searching Agris and Agricola, the two free databases available over the WWW would cover 367 of the primary agriculture journals. This would miss 175 primary agriculture journals, almost one-third of the primary agriculture journals. Combining CAB with Agricola or BAI, gains nothing since CAB covers both databases 100%. Combining CAB and Agris would search 502 primary titles. The Biosis and CAB combination would comprise of 506 journals, and the CSA and CAB combined search would include 501 titles. WoS combined with Agris searches 456 primary agriculture journals, and WoS with Agricola covers 428 titles. The combination of Agris and Biosis includes 404 journals, Agris and CSA comprises of 376 titles, and Agris and Agricola covers 367 journals. A combination search of Biosis and Agricola would include 365 primary agriculture journals. All other possible dual combinations exclude one-third or more of the primary agriculture journals. Very little is gained by combining three or more of the databases. The searcher will need to decide what level of exclusion of primary agriculture journals is tolerable when performing a comprehensive search.
2003 | Agricola | Agris | BAI | BIOSIS | CAB | CSA | WoS |
| Agricola | 232 (89.6%) | 104 (40.5%) | 205 (79.9%) | 259 (100%) | 159 (61.4%) | 233 (90%) | |
| Agris | 232 (68.2%) | 108 (31.8%) | 247 (72.6%) | 337 (99.1%) | 183 (53.8%) | 286 (84.1%) | |
| BAI | 104 (89.7%) | 108 (93.1%) | 103 (88.8%) | 116 (100%) | 91 (78.4%) | 113 (97.4%) | |
| Biosis | 205 (65.9%) | 247 (79.2%) | 103 (33.1%) | 304 (97.7%) | 194 (62.4%) | 283 (91%) | |
| CAB | 259 (51.9%) | 337 (67.5%) | 116 (23.2%) | 304 (60.9%) | 217 (43.5%) | 387 (77.6%) | |
| CSA | 159 (72.6%) | 183 (83.6%) | 91 (41.5%) | 194 (88.6%) | 217 (99.1%) | 199 (90.9%) | |
| WoS | 233 (58%) | 286 (71.1%) | 113 (28.1%) | 283 (70.4%) | 387 (96.3%) | 199 (49.5%) |
The overlap and percentage of coverage runs down each column of Table 3. CAB overlaps 100% of the Agricola and BAI journals. CAB also covers 99% of the Agris and CSA titles, almost 98% with Biosis journals, and the least with the WoS titles at 96% overlap. WoS overlap with the other databases is less since it starts with a smaller total percentage of the primary agriculture journals. CAB scans or indexes the most primary agriculture journals, so it has the least duplication with the other life sciences databases. When deciding upon which database to search, knowing the overlap among them may explain the result numbers the search produced and provide guidance if needing the most comprehensive search available. An opportunity for further research would be to start with the overlap of coverage of the primary agriculture journals, then compare citation overlap within the journals, or a more detailed comparison of subject/descriptor additions to the citations. This research would show which publishers add the greatest value to their database.
Discussions and Conclusions
This study indicates doing a search in CAB Abstracts for the agriculture journal literature will garner the most results. Doing an agriculture literature search in any or all of the other life sciences databases would be a duplication of effort. If one does not have CAB available for searching agriculture literature, then one must be knowledgeable in searching a number of databases to pull together a comprehensive search. These databases would cover the plant and soil sciences, forestry, human nutrition and health, engineering, food sciences, sociology, economics, animal sciences, education, and many other topics applied to the broad discipline of agriculture. Other databases to search might include Food Science and Technology Abstracts, Zoological Record, PubMed, Economics Literature, Engineering Index, and Chemical Abstracts. If one does not search CAB, an extensive knowledge of search tools, vocabulary, search interfaces, access to databases, and a vast amount of time would need to be available to the researcher.
Considering that BAI indexes or scans only 225 titles in all, and the description of this database states that this total is almost equally split between biology and agriculture disciplines with slightly more in biology, one can conclude that all of the agriculture titles in BAI are primary agriculture journals. This is a fine database to search when needing a small set of articles on an agricultural subject. Agricola covers many publications from the U.S. Dept. of Agriculture (USDA) which might be a consideration to the researcher. What Agricola does for the USDA is similar for what the Agris database does for the United Nations (UN). If a researcher is interested in UN's agriculture or rural information, one would want to consider searching Agris for the Food and Agriculture Organization and other UN publications.
For one to get the most comprehensive search both CAB and WoS must be searched. Combining the free agriculture database AGRIS with Biosis or WoS would provide the greatest gains if CAB is not available. If affordability is a concern, the two databases readily available for free on the WWW, AGRIS and Agricola would produce a decent search. The individual researcher would have to make the final judgment if the results were "good enough."
If the Library's Collection Development department is facing a budget reduction scenario, then evaluating the life sciences databases may help in the decision-making process. Of course, the information here is specifically related to the agriculture discipline. The databases compared are more general in coverage. CAB and Biosis are international and WoS focuses on all the life science disciplines, whereas, CSA seems to have the least benefit to the agriculture researcher. BAI offers a finite set of journals and Agris and Agricola are available on the WWW for free. The comparison of overlap among life sciences databases on primary agriculture journals may support a decision to terminate a database subscription if necessary. Further research on comparing citation overlap within journal titles or a detailed comparison of subject indexing added to citations would increase the knowledge of comprehensive agriculture literature searching.
References
Biological Abstracts, Inc. 2003. BIOSIS Serial Sources Philadelphia, PA: BIOSIS.Dote, G. & Letnes, L. 1996. Scholarly journals in agricultural economics libraries/reference rooms: a survey. Journal of Agricultural and Food Information 3(3): 47-63.
Hutchinson, B. & Greider, A. 2002. Using the Agricultural, Environmental, and Food Literature New York, NY: Marcel Dekker.
Kawasaki, J.L. 2002. Indexing of core agriculture serials. Quarterly Bulletin of IAALD 47(2): 33-37.
Kelsey, P. & Diamond, T. 2004. Establishing a core list of journals for forestry: a citation analysis. Issues in Science and Technology Librarianship [Online]. Winter (39): Available: http://www.istl.org/04-winter/abstract7.html [Accessed: June 14, 2004].
Murphy, P., Maddux, L. & Fenn. R. 1996. Current awareness services for agriculture: a comparison of serial title coverage. Journal of Agricultural and Food Information 3(4): 51-65.
Olsen, W.C. (ed.). 1991-1996. Literature of the Agricultural Sciences series Ithaca, NY: Cornell University Press.
Ulrich's Periodicals Directory. 2003. New Providence, NJ: R.R. Bowker.
Weintraub, I. 1997. Holistic literature searching for a holistic agriculture. Quarterly Bulletin of IAALD 42(1): 10-15.
