The fishing mortality (F) cannot be easily measured, at least over a short period. Control of F is therefore achieved by controlling some other quantity which is more easily measured and therefore controlled. The choice lies between catch (C) and fishing effort (f), e.g. total number of boats times the duration of fishing, which are related to fishing mortality by the two equations:
C = B x F ................ (1)
F = qf ...................... (2)
where B = abundance of the population and q = catchability coefficient
The implicit assumption made in controlling fishing mortality in terms of catch (or effort) is that B (or q) is constant, or that changes in B (or q) are known. Changes can be of two types - essentially random fluctuations, and trends. Particularly with long-lived species year to year fluctuations may not be very serious, even if not predictable. If the mortality in one year is high, because weak year-classes are present (for a fixed catch) this can be balanced by reduced mortality (lower catch or nominal effort) in the next, provided there is adequate administrative machinery to detect the occurrence and to make the necessary adjustments to the permitted level of catch or effort.
One major source of changes in population abundance - varying strength of year - classes - has been noted above. Typically a year - class enters the fishery when from one to five years old, and can make a substantial contribution to the catches for the next two to ten years. Forecasts of the abundance of the stock in one year can therefore be made from the information on the abundance and age composition of the commercial catch in the previous year, and these predictions will include all but the youngest fish. These can often also be predicted by research surveys of young fish, e.g. by the trawl surveys of young cod in the Barents Sea, and for long-lived fish quite accurate forecasts of abundance (excluding the effects of the removals by fishing) can therefore be made for several years in advance. The main practical problems are: first of providing the necessary machinery for rapid analysis of one season's results (catch rates, age composition, etc.) and subsequent adjustment of the permitted level of catch in subsequent seasons; and second, obtaining adequate estimates of the strength of year-classes among the youngest fish; if year-class strengths vary greatly and the fish are in the fishery for only a short time so that the strength of the year-class newly recruiting to the fishery can have a big effect on the total abundance, these latter estimates may have to be quite accurate and may require extensive surveys by research ships of the pre-recruit fish.
Even if natural (non-fishing) factors cause little change in abundance, a fixed catch limit can be very dangerous. The catch quotas set by the International Whaling Commission in the period after the Second World War (15,000 - 16,000 Blue Whale Units) were initially only just greater than the combined sustainable yield of the two main species (blue and fin whales). However, as the catches were too high, the stocks declined, and the nearly constant catch quota resulted in fishing mortality rates which exceeded the desirable rate to a greater and greater extent.
The control of the amount of fishing in terms of fishing effort would in principle prevent this happening. Thus, in the Antarctic, the number of catchers and the length of the season both increased during the period as the industry had to work harder to maintain the total catch. A limit on the number of catcher-days would thus have apparently kept the fishing mortality closer to the near-optimum level of 1946/8. However, even under the existing (catch-quota) regulations the power and efficiency of the catchers increased, and the enforcement of a constant number of catcher-days would have still permitted some rise in fishing mortality. In fact, it is very probable that regulation in terms of catcher-days would have in later years accelerated this process. Since 1962, there has been an agreement between the countries concerned to divide the overall quota. If this division had been in terms of effort, i.e. in numbers of vessels or of catcher-days, it would be greatly in the interests of each country to increase the catching power of its vessels, and hence increase the fishing mortality caused by its fixed share of nominal effort. The total fishing mortality would thus have increased beyond the desired limits. A better control of mortality could be achieved by the use of a rather more complex measure of effort, e.g. catcher days x gross tonnage of catcher, but it is unlikely that any simple and readily controlled measure of effort would take into account all the increase in catcher efficiency.
The control of mortality through effort control, e.g. in terms of number of days fishing, is subject also to other difficulties similar to those incurred when controlling mortality through catch, with additional difficulties due to the need for inter-calibration of effort statistics. The catchability coefficient, q, is no more likely to remain constant than the population abundance, so that measurement of fishing mortality in terms of fishing effort presents similar problems as the measurement in terms of catch. The variations in catchability coefficient, like the variation in population abundance, have to be measured, and, so far as possible, predicted. These variations have two groups of causes - natural and man-made. Generally natural fluctuations in catchability or "availability" are less violent than year-class fluctuations, but can be quite large. For example during cold winters in the North Sea soles can be much more easily caught than normally - an increase of several times in the fishing mortality per unit fishing effort; control in terms, say, of the number of hours fishing could allow serious over-fishing during a prolonged cold winter. Thus the winter of 1962/3 was doubly disrupting to the sole fishery - over supply during the winter, with a drastic fall in prices, followed by a shortage of fish in the summer, due to previous unusually heavy exploitation (and also unusually high natural mortality due directly to the cold).
Man-made changes in q can be larger, and will have more serious consequences on attempts to control mortality. Any statistics that can be readily collected concerning fishing effort will provide only an approximate measure of fishing mortality. For some gears, e.g. long-lines, an available unit of effort - e.g. the number of hooks times the number of sets made - may bear a fairly constant ratio to the mortality caused and improvements in fishing technique will be such as to increase the number of hooks set per boat or per man, rather than in the mortality caused by 100 hooks. For other more active gears, such as trawls or purse-seines (which take a major and increasing share of the world catch), the measure of both fishing power and fishing time (fishing effort equals fishing power multiplied by time) is less certain. For purse-seines (and increasingly for trawls, especially fished in mid-water), much of the time is spent searching for suitable concentrations of fish. Improved equipment (e.g. echo sounders, sonar) is continually increasing the effectiveness of an hour or a day spent searching.
Similarly the real fishing power of the average vessel in a fleet is being continually increased by building larger vessels, with more powerful engines, improved gear, etc. Some correction for this is possible by taking into account the various characteristics of ship and gear. This calibration is particularly useful in analyzing past data, especially when the introduction of new gears, or new ships, has taken place over a period. Thus several analyses have shown that, in a particular year, catches (and therefore the mortality caused) by trawlers of different sizes are proportional to both gross tonnage, and horse-power of the engines (Beverton and Holt 1957, Gulland 1956). This result is important when the average size, or power, of the ships in a fleet is increasing, but does not show whether these have been increasing in real fishing power of a vessel of a given size from one year to the next.
Such difficulties will exist in any fishery, but will be redoubled if the control of mortality involves any allocation. Then it will be greatly to the interest of each person receiving an allocation, presumably in the form of so many units of effort, to ensure that the mortality caused by his units (and hence the catch taken) is maximized.
The simplest case is a fisherman with a licence to operate one trawler; he will then build the biggest one feasible. This might be controlled by limiting the size to so many gross tons; then a bigger engine would be fitted. In turn the nominal horse-power might be controlled, but the engine run flat out (the extra catch making up for increased maintenance and replacement costs). Though the framer of the regulations may catch up with the ingenuity of the fisherman, it is more likely that the latter will keep one jump ahead. In any case, the vessel and gear and the methods of operations will be designed more to maximise the catch within the limits of nominal effort set by the regulations, than to maximise the catch from a given total of costs or to minimise the costs of taking a certain catch. To this extent the fishery will operate at less than its possible efficiency, and the magnitude of this inefficiency is likely to continually increase.
Changes in q, through improvements in gear, etc., will thus affect the total fishing mortality caused. If precise calibrations cannot be made, it may be possible to make reasonable forecasts of the likely increase in efficiency, q, for the fishery as a whole and take this into account when setting the limit to the total effort. For instance, if the efficiency is likely to increase at 3% per year, then the total allowed nominal effort should be reduced. However, in any complex fishery, involving several fleets using different gears, or different types of vessel, it is most unlikely that the improvements in gear changes in q, will be the same for all segments of the fishery. Therefore, the above adjustments, reducing the overall effort quota by, say, 3% per year to allow for the average increase in efficiency, while maintaining the fishing mortality at about the desired level, will not ensure that the shares taken by different segments of the fishery remain the same. This can be done only by determining the change in q separately for each group of vessels. In this, as in other matters, justice must not only be done, but also be seen to be done. For example if there are two fleets, trawlers (with effort controlled in terms of ton-days - tonnage of trawler times days at sea), and long-liners (number of hooks times frequency of use) it is almost inevitable that the long-line fishermen will suspect the trawlermen of achieving a greater increase in efficiency than is accounted for by any increase in tonnage, while the trawlermen may claim that their fishing power has gone up less than the tonnage. Conversely there will be suspicions that the fishing power of long-lines is not properly measured by the number of hooks. A related aspect of this is that the more precise measures of fishing effort rely on statistics of fishing time (e.g. number of hours spent fishing - with the net on the bottom in the case of trawls) which have to be provided by the fishermen (e.g. in log books) and cannot be easily checked independently. ("Black boxes" are being developed which, attached to an otter-board can record the depth and duration of each haul, but these are still a research tool). Also many of these log-books and other records provide one of the fundamental bases for the scientific study of the stock. If the form of the regulations encourages the falsification of the records, the research will suffer.
The basic check on whether two fleets have maintained the relative magnitudes of the fishing mortality caused by each is the ratio of the catches. For any year, or fishing season, the mean abundance B of equation (1) has a single value, applicable to all fleets, i.e. denoting the catches and mortalities caused by two sections of the fishery by subscripts.
C1 = B x F1
C2 = B x F2
whereas the corresponding forms of equation (2) are
F1 = q1 f1
F2 = q2 f2
where q1, q2 may have very different values, and vary in different ways from year to year.
For example, if in the base period of a management programme two fleets have taken equal catches, then they will have reduced the mortality caused by them to the same extent in some later year if, and only if, their catches are also equal in that year. There is an exception to this, if during the period taken as a unit for regulation (normally a year) there are big fluctuations in the population abundance (not availability) and in the time of operation of the two fisheries, so that it is not correct to take B as the same for both. An example of this is a gantlet - type fishery, e.g. of fish going upstream to spawn, with two fisheries at different positions in the migration route. In such a situation management practice is best formulated in terms not of fishing mortality but of the percentage of the stock starting on the migration. Then relative catches would again be the best measure of the relative percentages taken by two fisheries.
It appears therefore, that in a fishery in which several groups of fishermen participate, fishing mortality is best measured, and the targets of regulation set, in terms of catch. This is indeed what has been done in all the major international agreements concerned with limiting fishing on single species or groups of closely related species - c.f. Pacific halibut, Antarctic baleen whales, eastern Pacific tuna, Pacific salmon, etc. Only where a complex of species is concerned has some measure of fishing effort been used, c.f. controls on the number of vessels under some of the agreements concerning the bottom fisheries in the East China Sea.
The choice of the measure used to set a limit to the amount of fishing is only the first step; the second, and more difficult, is the choice of the control mechanism to ensure that this limit is not exceeded.