Cocks & Baird's reserve selections are based on a number of sets of initial
data, describing important characteristics of the available reserve sites
and the plant communities within them. Six types of plant community are
considered, which we shall name 
to 
. We use the predicate
to denote that C is a community of interest in the reserve
selection process. Thus, to state that 
is such a community, we
write 
. We do the same for the other five communities.
There are 101 potential sites which we might use as reserves and,
for simplicity, we shall use the integers from 1 to 101 to refer to
these sites. For every site Cocks & Baird provide a simple numerical
measure of abundance for each community. We use the predicate
to denote that the index of abundance of community
C at site S has the value P. This can vary from 0 (no chance
of appearing in some standard sample) to 1 (always appearing in some
standard sample). The clauses
corresponding to this information for all communities in site 1 are:
Similarly, we create appearance clauses for each
community on each of the remaining 100 sites. Each site also has a known
area which we record using the predicate 
, where S is a site and
A is its area (in 
). Thus, for site 1 the area is defined as
.
It is important to know whether sites possess inland and/or coastal
habitat. We use the predicates 
and 
to denote these
properties, respectively, for any site, S, to which they apply. For
example, we know that site 10 is inland (
) and site 1 is
coastal (
).
Finally, a few of the sites are already assigned as reserves. We use the
predicate 
to denote that a site, S, has this
property. For instance, site 99 is an existing reserve so we define
.
This completes our description of the site data given in Cocks & Baird's paper. To save space, we have not given the complete set of definitions for each of these predicates but it is straightforward to reconstruct these from the original paper, following the procedures given above.