Interest is Paid on Deposits in your Provincial Data Bank
ADDRESS:Eastern Ontario Biodiversity Museum, Box 1860, Kemptville, Ontario K0G 1J0 - e-mail: email@example.com
Database structure table
As a peer-reviewer of herpetological manuscripts, I'm frequently struck by how many authors fail to deposit copies of their data in a publicly accessible archive. Not making data public has such a long history that it has become one of the major structural weaknesses of observation-based natural history, bleeding away masses of painfully acquired information that could have been invaluable for replication of results or reanalysis but are not (editors' "cannot be") presented in published papers.
The traditional archive for natural history work was the museum collection. In many institutions, especially in the American West, this was supplemented by the Grinnell System of field notes, deposited with the specimens in a public museum. The Grinnell System used a standard format of Journal, Species Accounts, & Catalogue, on archivally stable materials bound into volumes. It was devised in the late 19th century to ensure that field observations were permanently associated with collections (Herman, Steven G. 1986. The Naturalist's Field Journal: A manual of instruction based on a system established by Joseph Grinnell. Buteo Books, Vermillion, South Dakota. 200 pp.). In the 20th century, some authors deposited hard copies of their data in libraries, but retrieving these documents is slow and uncertain. On the other hand, many scientific disciplines have now set up internet-accessible data banks, the most prominent, perhaps, being the sequence databanks of molecular geneticists (see http://www.ncbi.nlm.nih.gov/Genbank/index.html ), and have made deposition of data in such an archive a prerequisite for publication.
Since the conclusions of natural history are not the robust universal laws of the easy sciences, or the repeatable experiments of the indoor sciences, but rather tend to be fragile interpretations of masses of contingent out-of-doors observations, it is especially important that we have means of archiving the data on which our conclusions are based. Beyond ‘confirming' that we actually made the observations behind published conclusions, archives allow others to reanalyse them, either to check our calculations or to test new hypotheses against the data. Archives also facilitate generalization across studies, allow repetition of observations in future decades when conditions may have changed, and have many other uses. Many of our data are just records of the presence of a particular number of individuals of a species, engaged in a particular behaviour, at a particular time and place, and may bear on a wide range of hypotheses. Lamont Cole contended (in the 1960's) that ecology was mostly the science of the occurrence of organisms at particular times and places, and while this now understates the wide range of phenomena natural historians study, it is still true that ‘occurrence' reflects all the factors acting on an animal, not just the ones of interest in a particular study, and the data of occurrence can be useful from many points of view.
‘Occurrence' is what atlassing projects seek to capture, and it is because so many of our data take the form ‘species/ time/ place/ habitat/ weather/ (behaviour/ size/ notes-peculiar-to-the-study)' that atlassing schemes (and here my experience has been with the Ontario Herpetofaunal Summary or OHS) can provide an adequate framework for archiving data. At least in Ontario, the OHS has become the de facto repository for all amateur, public, and government herpetological observations, as well as (thanks to the industry of Wayne Weller and Mike Oldham) an index to museum collections from Ontario and to the entire published & unpublished herpetological literature of the province. The data of many academic herpetologists are conspicuous by their absence. Atlassing schemes often have a fairly restrictive format, but a little stuffing and prodding by the investigator should make it possible to fit the essence of any observation into an atlassing format, so that someone using the published paper as a guide would be able to extract those data that are not usually included in an atlassing record.
This is not to suggest that the format used by atlassing projects is ideal, or even realistic for modern computers. The OHS format was originally constrained to fit onto one line of a wide-paper dot-matrix printer, a criterion which is certainly obsolete today. Table 1 lists the structure of the main table of our database at the Eastern Ontario Biodiversity Museum, which is descended from a combination of the fields of the OHS database and of the National Museum of Canada herpetology catalogue card, and might contribute to discussion about a mature structure for atlassing databases.
It would be nice to have a central national or continental database to hold all the records gathered by regional databases, but there's no obvious Canadian host for such a database. Before 1990 the card file of the National Museum of Natural Sciences tried to serve as such a database, but with the withdrawal of the now-Canadian Museum of Nature from herepetology it hasn't got the interest or personnel to host a national database. EMAN encourages such databases, but doesn't host them, and CARCNET's funding is too meagre to provide long-term security. The provincial Conservation Data Centres like Ontario's NHIC (which now holds the OHS database), are widely involved in provincial inventory and atlas projects. One of their mandates is as "central repositories" for natural history data in their jurisdiction, and as such they are natural hosts or co-hosts for the provincial databases.
Even without a national repository, I suggest: 1) That authors, editors, and referees make deposition of data in a provincial or regional database a prerequisite for publication, 2) that ‘atlassing schemes' formally acknowledge the larger role they are playing, or can play, by explicitly becoming permanent, precisely geo-temporally referenced, historical general archives, rather than episodic ‘1-record/ quadrat/ decade' mapping schemes, and 3) that ‘atlassing schemes' upgrade their database structures to handle a wider range of information, including ‘negative data' of searches or auditions where a taxon was sought but not found, finer geo-temporal referencing, and memo fields in which any sort of text, image, or numerical data can be stored.
TABLE 1. Structure for a General Natural History Database.
This is the structure of the main table of the Eastern Ontario Biodiversity Museum database of observations and collections, with ‘lowercase' spellings-out of the ‘UPPER CASE' variable names. A *.dbf file in Foxpro 2. from: Frederick W. Schueler & Anita Miles, 2000, Establishing EOBase, a Database of Eastern Ontario Natural History Collections & Observations, at the Eastern Ontario Biodiversity Museum. unpublished report to the Eastern Ontario Model Forest. 97 pp.
Field Name || Type - Width /Decimals |
| 1 sourceTABLE || Character 12 |
| 2 ENTry_NUMber || Character 30 |
| 3 EVENT* || Character 31 |
| 4 LAST_MODified || Character 25 |
| 5 FIELD_NUMber || Character 25 |
| 6 SPECimen_NUMber || Character 25 |
| 7 SOURCE || Character 40 |
| 8 OBSERVER || Character 60 |
| 9 NAME || Character 50 |
| 10 SUBNAME || Character 50 |
| 11 CLASS || Character 10 |
| 12 DETermined_BY || Character 30 |
| 13 LIFE_STAGE || Character 20 |
| 14 HOW_CAUGHT || Character 20 |
| 15 COUNT || Character 30 |
| 16 DISTrict || Character 15 |
| 17 TOWNSHIP || Character 20 |
| 18 LOCATION || Character 50 |
| 19 topo MAP || Character 20 |
| 20 utm 100km SQUARE || Character 7 |
| 21 EASTing || Numeric 10 1|
| 22 NORTHing || Numeric 10 1|
| 23 LATITUDE || Numeric 9 5|
| 24 LONGITUDE || Numeric 10 5|
| 25 COORDinates || Character 30 |
| 26 ACCURACY || Character 30 |
| 27 ELEVATION || Numeric 6 |
| 28 HABITAT || Character 80 |
| 29 AIR_TEMP || Character 20 |
| 30 TIME || Character 20 |
| 31 DAY || Character 7 |
| 32 TIME_ZONE || Character 10 |
| 33 MONth || Character 5 |
| 34 MOnth || Numeric 2 |
| 35 YEAR || Character 9 |
| 36 REMARKS || Character 57 |
| 37 TEXT || Memo 10 |
| 38 MEASUREments || Memo 10 |
| 39 CURATION || Memo 10 |
| 40 HISTORY** || Memo 10 |
* This identifies all the records resulting from one observation event or vist to a site.
** ..of the manipulations of the record within the database.
SCHUELER and KARSTAD PUBLICATIONS
skip to top - skip to CONTENTS back to publications page - Pinicola home page
COMMENTS: This is my plea for universal public archiving of herpetological observations. There has been some talk since (CARCNET 2001 in Charlottetown) about a national database, and a uniform format, for herpetological obsersrvations, but no funding or action. F.W.Schueler - March 2003.