Specimen databases

[Entomology: Namibian Museum]

For what it is worth, I would recommend careful consideration of Christian
Thomson's comments. Our Museum is not in the same league as was most
institutional responses to this string. We cannot afford a full-time
database manager, cannot develop our own dedicated software, cannot maintain
databases requiring expensive networks, yet wished to make available
information on our collections. For us a major consideration the purpose of
a collection database, but also cost. Our goal was to know what is
available, from where and when, in order to have a more considered approach
to project planning. Such information would allow us and prospective users
to rapidly evaluate collections to identify particular needs. However, if
the goal of a database is research, employment, or to access funding, then a
whole different scenario would apply (-:

I have yet to hear of one institution where comprehensive, retrospective
databasing has lead to a substantial saving in personnel or administrative
costs. In fact, it usually just added another financial burden which impact
on more crucial expenses - a case of the tail wagging the dog. Our project
parameters were to keep expenses as low as possible, use the minimum of
staff time, get the maximum number of specimens onto a database in the
shortest possible time, and maintain and improve databases cost-effectively.
Such simple requirements have little need for many of the whistles and bells
in databases. The ASC model, for example, includes a large number of
unnecessary wish-to-have's which scientists are not going to use. Most would
examine the specimens themselves rather than trust remote data inputs, even
if done by peers.

I therefore examined various applications like BIOTA, SPECIFY, BioLINK, and
others. We considered data entry (time cost) and staff training for using
the databases. In principle they were all very similar, but, for what it is
worth, Biota was most appropriate, particularly because it is a low-cost
cross-platform application (both Mac and Windows versions), well tested, and
with an excellent hard-copy manual. Yes it has its problems, but so do most
other applications. Our major problem was actually to have a simple and
effective data capture system. We designed that ourselves, for which we used
FileMaker Pro for its simplicity and because we had the user experience.
However, any of the other commercial database applications may have
sufficed, and I've often seen spreadsheets used for that purpose too. Our
input application allows a single datatypist to capture up to 6000 records
per month (14 fields), offers pick-lists to restrict errors, and allows
rapid verification. As an example, a local custom-designed avian database
with some 25000 records was established for US$ 70 000 over 2 years. We have
completed a similar-sized mammal database, including evaluation costs, at a
cost of US$ 2500 over 2 months. Once records have been verified, additional
fields may be added rapidly (higher classification, expanded provenance
information, specimen management information, etc.). At that stage the
initial database can also be exported to a more comprehensive collection
management database.

Our big challenge is still entomology, primarily because I cannot afford
either the staff time or other resources for a low-priority product. Other
collections within our Museum had different priorities. For entomology we
have developed a novel approach to data entry, though inherently it has huge
verification problems. The corporate community is sponsoring internet and
computers at schools through a competitive environment. The competition is
to transfer paper-based catalogues to database - see
http://www.natmus.cul.na/insectathon.html - and part of the verification
solution we are now testing is to duplicate entries, and then to
electronically compare individual records for errors. For all participants
it is a win-win situation, we may obtain a useful product for relatively
small investment, the corporate community raise their public profile, and
the schools gain valuable experience and resources.

My advice to anyone considering databasing their collections would therefore
be to determine the particular purpose of the database and how much
resources, immediate and long-term, can be allocated. Then evaluate which
applications are available and enquire for experiences from others.
Practical evaluation is time well spent as it is possible to know the
limitations of particular applications within a week or so. Then prioritize
and think-tank your project institutionally according to standard
project-design principles. Only then decide how to proceed.

Regards
Eugene Marais

Entomology Centre
National Museum of Namibia
P.O. Box 1203, Windhoek, Namibia
http://www.natmus.cul.na/newindex.html