Íslenskar landbúnaðarrannsóknir - 01.09.1978, Qupperneq 47

FISHERY STATISTICS 45
transition. These data are used for
generating the primary íile (SALMAF).
One record in this íile represents one entry
in the logbook, usually one fish. During
the initial stages of the computer process-
ing, this primary file is condensed into a
secondary file, consisting of five arrays
(Table 1), which is the data base on which
the program operates in producing the
output. This secondary file can also be
stored on disk for later processing. »
DATA PROCESSING
The data processing can be done in three
ways (Fig. 2).
1. Single primary file.
A single river processed as a whole, where
the input is one primary file. A secondary
file is formed, stored permanently on disk,
and processed to give the printed output.
2. Multiple primary file.
Many primary files form the input, usu-
ally many sections of the same river. A
secondary file is formed for each, stored
temporarily on disk, and processed to give
a printed output for each section. Finally,
the stored secondary files are combined
into one and processed for a combined
output for the whole river. This final sec-
ondary file is stored permanently on disk.
3. Secondary file.
If the user wishes to combine rivers which
are already stored permanently on disk
into secondary file form, he may retrieve
any number of secondary files and com-
bine these for processing.
Output.
The printed output lists the various
statistics in tabular form. These include a
summary of totals by species, sex and bait,
total and average weight by species and
sex; and a weight distribution for each
species. Also listed is information on daily
catch, expressed as percentages, as well as
totals of the cumulative catches (Fig. 3).
DISCUSSION
The data file is a method of storing catch
data for a salmon fishery in an easily re-
trievable form. The file is flexible in that
additional information can be added
quickly and the programs can be modified
easily or expanded since there is a main
program which calls up short subroutines,
each of which can be changed separately.
The card file (primary) eventually can be
written on tape for convenient storage.
In the future, when the daily run size
can be estimated independently by the
electronic counters present in a few rivers,
it is anticipated that extensive analysis
will be required to relate functionally
catch per unit of effort to the daily counts.
The catch file will then serve as the prim-
ary data base.
The principal cost of maintaining this
file is the cost of card punching. With an
average of 30 to 40 thousands cards (fish)
per year, the present cost is 300 to 400
thousand Icelandic kroners or about U. S.
$0.05 per fish. The computer processing
itself is short and therefore inexpensive;
only the input/output requires extensive
time. Modifications of programs or de-
velopment of new programs represent ad-
ditional expenses.