Standards for zoological databases
Standards for zoological databases were elaborated as a result of ZOOINT
project development.
Some main positions, rules and standards for construction of biological
databases:
-
The relational databases model should be used.
-
According to requirements of databases normalization, information
systems should include separate tables on each aspect of the information
about animals or plants.
-
Each information unit should be stored in the specific system only in a
single record of a single table. This rule requires the wide usage of
dictionaries and classifiers (or thesauri - dictionaries with a hierarchy
structure and generic-specific relat ions between concepts).
-
Classifiers of animals, plants, mushrooms, bacteria and other kingdoms
of living organisms should lay in a basis of the system.
-
The data on geographical distribution of organisms, frequently used by
biologists, should be strictly divided into separate aspects
(administrative-territorial, physiographic, biogeographic, landscape,
natural-zonal, etc.). Terms of each aspect should be fixed in
classifiers.
-
It is impossible to achieve a support for different databases or
databanks in one DBMS even in one institute and it is not necessary to aim
at this. However, it is important to use professional commercial DBMS
with approved import and export mechanisms . As a criterion of DBMS
applicability it is proposed to use DBF (dBaseIII) format. DBMS not
capable to import or export files from or to this format should not be
used.
-
In order to simplify data exchange between DBMS working under different
operating systems, it is necessary to limit the length of filenames with
8, extensions - with 3, and fieldnames of databases - with 10 characters;
besides only Latin letters, digit s and underline symbols with no
difference between upper and lower case (if a system supports such
possibility) should be used. Field length should be less than 255
characters in systems, where character field length is not limited or can
be very large.
-
Nevertheless, if the hierarchy is represented without the use of ZOOCOD
standard, it is necessary to aim at some rules simplifying the transition
to this standard:
- a) the use of the only field of the table for indication of a taxon rank
or concept with stored codes or complete rank names;
-
b) there should be a field with the unique code of a taxon and these codes
should be used for the link of taxa tables with data tables;
-
c) separate field should contain information on synonymy with unique code
of valid name for each synonym;
-
d) another field should be used for indication of the systematic order of
records (ordering on this field should create an analogue of the
"systematic list" which is common and necessary for the majority of
biologists);
-
e) two fields should reflect hierarchy in a way, accepted now in Windows
applications (fields for presentation of hierarchy by an ASP method
"TreeView"). This does not cancel the necessity of items a) and d) for
these two fields do not reflect true ranks of taxa and required order of
one parent daughter's taxa.
Authors of this page:
Andrei L. LOBANOV
Igor S. SMIRNOV