This is a popularising account of a project done by the writer in 1974-5 when with the TCD Department of Statistics, with some help from Aoileann Nic Gearalt of the Tralee Regional Technical College, who had earlier completed an MSc in TCD under the writer's supervision, on a somewhat related theme. The project was sponsored by Gaeltarra Eireann, the Gaeltacht development agency, and the report was published by Arthur Reynolds in the May 1975 issue of the Skipper, the trade journal of the fishing industry. The report was extracted by Arthur Reynolds from a longer article which had appeared first in the IMI publication "Management".

A Study of Fishing Port Efficiency

Quite independently of this, we had been exploring, with others, how the art of systems modelling could be applied to problems of regional development in the west of the type encountered by Gaeltarra Eireann. After about six months of interaction in the earlier part of 1973 we came up with a project to try and quantify the specific development problems faced by the smaller Western ports, such as Dingle, Rossaveal and Ballyglass.

The latter is of particular interest because it is close to the rich fishing grounds of north-west Mayo, and is in an impoverished area.

We discussed this with Alex Heskin of BIM, and it was complementary to the national model planned by BIM with Dr. Lennon. An output from the Gaeltarra model, quantifying the flow of fish from a future developed Western port, would constitute in added input for the BIM model. An input to the Gaeltarra model (fleet size and composition) could emerge as an output from the BIM model, or it might need to be assumed as input.

In other words, we were constructing a co-ordinated system of planning models.

Vicious Cycle

The market is unreliable because of the lack of shore-processing, and the remoteness of the fresh market. Harbours are undeveloped because the local authorities are not entrepreneurs. They act in the interests of actual rather than potential rate-payers.

Shore-processing is undeveloped because of lack of continuity in the supply of fish. The problem is: where do you begin, and how do you predict the pay-off ?

There is a body of experience emerging in the Trinity College. Dublin Statistics and Operations Research Laboratory with regard to problems like this.

DINGLE SELECTED

We also established relations with Burtonport, Co Donegal, visiting the co-op there on a number of occasions. Broadly speaking, we developed the model on the basis of !the experience of Dingle, and then put in the factors appropriate to Burtonport to see if we could predict the performance of the Burtonport system in a manner which fitted experience. This was a rudimentary validation procedure.

We are indebted to Pat Brosnan, secretary of the Dingle co-op and to Pat Bonner, secretary of the Burtonport co-op, for invaluable assistance, especially in the case of Dingle where the data collection occupied co-op stall time over the whole season of 1973.

The analytical system consists of three parts:

(a) A Harbour. model which predicts the amount of available weighted fishing time as a function of tidal constraints, berthage, fleet size anti composition ("weighted" time because an hour of fishing gives a catch which depends on boat size).

We build into this a measure of social conventions in the form of an average working day over the tide-cycle, and the maximum day. Distance from fishing-ground is also relevant input;

(b) A statistical data-gathering procedure which relates to the performance of a particular fleet on a particular fishing-ground; this gives a factor converting fishing hours into tons per hour by boat-class by species in any month of the year;

(It is at this point that inputs from marine biology would be relevant, if we had them. There are mathematical models of fish populations which are in use for predicting the response of a fish stock to intensified exploitation. We did not develop this angle on the problem, as data is lacking. This is another research programme, which should land on the agenda of the Galway marine research people. Are we dealing with relatively unexploited stocks, or have illegal foreign trawlers already eroded them to the diminishing returns level ? Whichever is true, it can be assumed that an extra effort by Irish-based fishermen will bring in extra catch in some proportion to effective fishing time; in the latter case the proportion might be less than equal). (c) A shore-processing model, which allocates the catch to one or more channels (these can include sales on a remote fresh market) taking into account current prices, processing costs and availability of a given investment in processing capacity. The output appears as a cumulative cash flow and a production plan. Fixed costs are allocated appropriately to each line and taken into account in the costing. The unit-costs therefore have a utilisation-dependent component and are a function of seasonality of supply. Labour demand (both permanent and temporary) is also estimated.

NEW PROPOSITION

Assume we have some data, collected as in (b) above. Step one is then to try out a plausible shore-processing system with the present catch figures and test for viability. if non-viable, increase theoretically the catch until the viability threshold is exceeded. This gives the measure of improvement of the harbour-fleet system that is necessary.

Step two is to expand the fleet (theoretically) so as to, supply the required flow of catch. A run at the harbour model with the expanded fleet will show up the bottlenecks if any. Further runs with an expanded fleet, improved access and/or extended berthage will theoretically remove the bottlenecks to an extent which can be calculated, giving an estimated performance for the improved harbour and expanded fleet in terms of weighted fishing hours.

The costing of this is usually a matter of civil engineering and is known to the Board of Works. Whether this expenditure is economically justified can be estimated using the cash-flow projections produced by the shore-processing model. it is necessary to allow separately for capital invested in boats, of course, as the expanded fleet does not appear by magic.

Thus we have developed an analytical procedure which enables a port investment programme to be systematically evaluated, and, what is perhaps more significant, alternative development programmes compared.

Unlike the BIM national model, which is running on a routine basis, we have not yet had the chance to carry out a definite port study. One problem is whom do we do it for? Who pays? If fleet harbour and shore processing were under an integrated management (analogous to the farmers with their bulk collection systems and food processing factories) there would be no problem. For as long as individual fishermen confront a single privately-owned shore-processor, with fresh-market oriented merchants in competition, it will be difficult to plan.

It would be possible for a local authority to use the harbour model to help it make decisions, or a processor to use the shore-processing model, but any plans evolved could be negatived by someone else's decision. A processor who owned the harbour and the fleet and, employed the fishermen could, of course, benefit by investing money in this type of techno-economic modelling approach to planning. This is basically the pattern in the developed part of the industry as it exists in Grimsby etc, with fleets of large boats fishing in remote waters.

The Irish fishing industry, however, is not yet at that stage; nor need It develop unduly in that direction for as long as there are rich fishing-grounds within an hour or two's journey from a Western port.

CO-ORDINATION

For example, a port development committee, under Gaeltarra chairmanship, consisting of fishermen's co-operative representatives, local authority representatives, processors, etc., meeting round the table, could go through a planning exercise, during the course of which a number of "what if . . ." questions would come up.

It would be the function of the writers to attempt to answer these questions quantitatively; if the computer model in fact contains all the essential features.of the situation it would be possible sometimes to have an answer the next day, or at worst within a week or two.

Some navigational notes:

Copyright Dr Roy Johnston 1999