2.3.1 Statistics of the structure and operation of the primary phase
2.3.2 Statistics of the secondary and tertiary phases
Fisheries scientists have filled a key role in formulating the objective, of management and hence in setting up the organizations and systems for trying to achieve them. At present they carry a very large responsibility for ensuring the clarity and credibility of their advice, particularly in matters of international regulation. This means that they also carry a large part of the responsibility for ensuring that the basic data requirements are being met and that the case for their collection is presented clearly.
A recent working party of the Advisory Committee on Marine Resources Research (ACMRR) discussed the qualities required for scientific advice, including timeliness, accuracy and precision, clarity and ease of comprehension, scope and relevance,
Figure 6: Institutions responsible for collecting, analysing and acting on fisheries statistics
A. Functions all performed within one ministry
B. Functions spread over several institutions
acceptability and credibility (ACMRR(FAO), 1974). The same qualities apply to data, and it is very often failure in the supply of data which lead to inadequacies in the scientific advice. It is fairly easy to appreciate from a common sense point of view why these qualities are important if we relate them to a simple example like the one given in the introduction.
A census of population to provide information on the planning of hospital and school building over the next five years will only be useful if the results are available fairly quickly and in a form which the planners understand and trust.
Difficulties arise for a variety of reasons, which can be rectified only if they are made clear and the will to rectify them exists. Perhaps the greatest of these is that the collection and study of data, formulation of advice on specific problems and actual planning and implementation of a management policy are carried out in a number of stages and often by different people or groups of people with different interests. Even if they all work within the same ministry or organization and the objectives of the collection of statistics are laid down, it is vital that the purpose of each step in the process is clearly understood by those undertaking the work. In practice it is more usual to find that the data collection system serves a variety of users and that, for example, the failure to produce a particular set of statistics on time is due not to laziness or incompetence, but to the lack of awareness by all those concerned of the overall functioning of the system.
Although very difficult to overcome once a system has become institutionalised, many of these problems could be avoided by clear planning at the outset and intelligent use of management techniques. The kind of system analysis and flow charting needed in order to implement systems of automatic data processing will also often help to reveal bottlenecks and inadequacies. Automatic data processing itself may provide a cure to some problems of data handling, but only if the necessary clarity in defining objectives has been achieved. Regular reappraisal is needed with all collecting and processing systems, irrespective of their level of sophistication.
As well as these institutional problems it is absolutely essential that the data should be defined clearly and unambiguously in all classifications and tabulations, that totals arrived at by different routes should correspond, and that figures which appear in different places (e.g., exports from country A to country B, imports by country B from country A) should be the same or that any discrepancies should be explained. The prevention of such discrepancies and the use of checks in processing will be dealt with in Section 4.
The term "catch" is often used loosely, when what is really meant is "landings". "Landings" should always be used to mean the actual weight of the fish landed, whether it is gutted, filleted., frozen, reduced to meal and oil etc. "Catch" or "nominal catch" is the live weight equivalent of the "landings". Factors for converting "landing to "nominal catch" should be calculated for each species or group of species for each way of landing, e.g., frozen whole, frozen filleted, unfrozen gutted. There may even be variations for the same species from different areas caught at different times of year which have to be taken into account. For scientific purposes, international statistics of fish catches should be of "nominal catch". Since the biologist is concerned with the effect of fishing on the stock rather than with the supply of fish at the port he also needs to know the quantity of fish discarded at sea plus any other losses between catching and landing. Taken al together these make up the total weight of fish
Fig 7: The definitions of catch
Fig. 8: A sample of catch and effort statistics compiled in the U.K.
removed from the sea, or "gross catch". In some instances the proportion of the "gross catch" which is discarded is very high and since the fish involved are usually small it may represent a very high mortality on young fish. Figure 7 shows the relationship between the various catch definite ions. Notes on conversion factors, time periods, direct foreign landings and classifications into specie are given by FAO, Department of Fisheries, Fishery Economics and Institutions Division (1973). These definitions and classifications must be strictly observed in order to en sure that statistics collected in different places or at different times are comparable.
The degree of subdivision of the catch data by species, area, time period, vessel type etc. will of course depend on the use to which the data are put. A balance must be struck between the degree of subdivision and the cost of collecting and processing. (For example, to assess the effect of proposed. gravel dredging on a small trawl fishery one needs detailed. information on the quantities and values caught near gravel deposits in the area). Since the number of possible species x area x time x vessel type x gear x port combination is enormous it is impossible to produce them all and instead a, small number of the more widely used ones are tabulated. Others can be obtained as required. With a full data storage and retrieval system it should be fairly straightforward to store the original data and reprocess it in different ways when needed. Table 3 gives a list of some of the tabulations of catch statistics for demersal fish available in the U.K. with an indication of those which are most commonly used. An example of one of the tabulation., which are output by a computer line printer, is given in Figure 8.
The major publications of international fishery statistics give catch by species or species group by year for each major fishing area or for sub-areas within them. The Coordinating Working Party on Atlantic Fishery Statistics holds regular meetings to review categories and classifications used and to improve the methods of compilation, processing and dissemination. The report of the 8th session (FAO/ICES/ICNAF/ICCAT/ICSEAF, 1974) deals among other things with the most recent updating of area codes in the Atlantic, classification and codification of species items worldwide and the International Standard Statistical Classification of Fishing Vessels. FAO, Department of Fisheries (1973) provides more dc tails of vessel and gear types and fishing areas. Classifications of species and species groups for statistical purposes are being continuously updated (e.g., FAO, 1974). where certain landings or parts of landings cannot be assigned to a species or species grouping they may be classified as "other", but any information about the species which are likely to be lumped, together under this heading should be recorded and reported.
We have already stated that in general terms fishing effort is the work done in catching fish, i.e., the input of labour, vessels, skill and technology, but that it does not have a standard physical dimension (e.g., joules). The classifications of areas, ports, vessel categories and time periods used for recording effort should be the same as those used for recording catch so that the effort used for a particular catch is known. The economist is concerned to translate these inputs into terms of money, taking account of the opportunity costs. The biologist is not concerned with the money value or even necessarily with the physical values of the inputs as such. For him one unit of fishing effort (f) removes a constant proportion of the stock and is directly related to the fishing mortality (F) by a constant known as the catch-ability coefficient (q), i.e., F = q.f. For example, if the fish in a pond are evenly distributed over the bottom and one takes every fish in 1/4 of the area then one takes 1/4 of the stock regardless of whether there are 100 fish or 1 000 fish in the pond. The numbers of fish taken by one unit of effort, in this case 25 or 250, (= the catch per unit of effort) are art index of the difference in abundance.
Table 3: Some of the tabulation, of catch and effort statistics of demersal fish available in the U.K.
# The tabulation marked # is showen in Figure 8
+ Mark those. Frequently used
In practice the use of fishing effort as a measure of fishing mortality or of catch-per-effort as a measure of abundance (given total catch the two are essentially the same thing) is far more difficult than the simple example would suggest for two reasons (1) the catch-. Ability coefficient q is not constant because in a real situation the fish are not evenly or randomly distributed and neither are the fishing boats; (2) what we are able to measure directly is not the proportion of the stock caught, which is our definition of fishing effort, but some or all of the factors which we think may affect the proportion of the stock caught by a vessel or fleet in a particular time period.
Taking the first of these, let us suppose that the fish all come together for spawning at one time of year. At this time one unit of physical input (e.g., one day's fishing) in the spawning area will catch a far higher proportion of the stock and cause far greater mortality than when the fish are scattered. However, if such seasonal changes are a regular annual feature then the catch-per-effort in, for example, a spawning fishery will still be an index of year to year changes in abundance. If such a year to year change in abundance is all that is needed from the effort measurement then it may be sufficient to collect this datum only for a short period each year, during which it can be well and easily estimated. We are thus allowing for a change in the catchability coefficient, q, caused by the regular seasonal migrations of the fish. Short term fluctuations in catchability caused by, for example, diurnal migrations can also be allowed for, but often it is necessary to assume, as the simplest hypothesis, that q has remained constant, in the absence of any evidence to the contrary.
The second difficulty, that of finding an index of fishing effort, has been the subject of a very large body of research (e.g., Pope, 1975). Rothschild (1972) has suggested that the factors which affect fishing effort (e.g., the product of time spent fishing and gear, horsepower and size of vessel, etc.) should be called "fishing inputs". They are also sometimes known as "nominal effort". Fishing effort can be regarded as the product of fishing time and fishing power, which in relation to a uniform density of fish are defined as follows: the fishing power of a vessel or fleet is the quantity of fish which it catches per unit of time, relative to other vessels or fleets, and the fishing time is the time during which catching takes place. In other words if two boats fish side by side for an hour and one catches twice as much as the other then its fishing power is twice that of the other.
The quantity of fish which a vessel catches per unit of time (i.e., its fishing power) will be determined primarily by the size of the vessel, its horsepower and the type and size of gear used, but a large number of other factors can play a part. Among those which have been investigated are the age, storage capacity and method of construction of the vessel; the size and skill of the crew; the use of technological aids such as Decca navigators, echo sounders, sonar, power blocks and in many oases, the particular species sought. Some of these factors cannot be physically measured and the ways in which they affect fishing power are complex and liable to change; nevertheless they provide a framework for dividing a fleet into categories within which fishing power is less variable, and these categories from the basis for the classifications of fishing effort given by FAO, Department of Fisheries (1973).
For each effort category (or more strictly fishing power category) we need a measure of the fishing time. For some gears (e.g., trawls), the quantity of fish caught will depend fairly closely on the time that the gear is actually fishing, and hours fished or number of hauls, if the haul length is constant, will be a good index. The time in the water will not be a good index for gears which become saturated (e.g., a long line on which all the hooks are taken or a drift net with clogged meshes) or for gear in which searching for fish plays an important part (e.g., whaling, purse seining).When deciding on the measure of fishing time appropriate to a particular situation the question to be asked is "would we expect the catch to go up in direct proportion to this time unit, given that fishing power stays the same?" The measures of fishing time used in most fisheries include: number of hours fished; numbers of hauls, drags or sets made; number of days fished.; number of days on ground; number of days absent from port; number of trips made. Since these measures are not mutually exclusive they can all be measured for each vessel if desired and the moat appropriate one selected. For a purse seine fishery it may be useful to know in. addition the time spent handling the gear, the steaming time to and from the grounds and the searching time on the grounds. Even in the situation where the statistics of fishing power and, fishing time available are detailed and appropriate, the biologist will require great skill and knowledge to interpret them in constructing his index of fishing effort.
If no conventional statistics of fishing effort as listed above are available it may be necessary to look elsewhere for data. One example of this is the use of records of fuel consumption, kept by the tax authorities, as a measure of the work done by fishing boats (Levi and Giannetti 1973). This is a rather direct measure of work, but if, over a period, relatively more fuel on a boat is used for refrigeration and ancillary equipment than for fishing, this will give the appearance of an increase in fishing effort. This is also the case with other fishing effort measures used to compare freezer trawlers with fresh-fish trawlers. A fresher has a higher fishing power than a freezer of the same size and engine power and their fishing tactics differ, since a fresh-fish trawler is concerned to fill up as fast as possible, while a freezer has to work more steadily to stay within the capacity of its processing equipment.
In choosing statistics for measuring "fishing effort" or "fishing inputs" for whatever purpose, there are two factors which must always be kept in mind.:
1. The statistic chosen must be widespread, persistent in time, easy to measure, capable of aggregation within each effort category and unambiguous.
2. The concept of "fishing effort" implies maximization of some output (i.e., the work done), whether explicit or implicit, on the part of the deployer, and there must be some idea, however vague, of what is being maximised..
Catch per effort is usually found by dividing the catch for a particular area,/port/vessel category/time unit by the equivalent effort. In many cases only the effort for a small group of vessels within the vessel category or for one vessel category out of the many landings within the aria/port/time unit is known. Provided the equivalent catch is available it is then possible to rise up this partial effort figure by the total catch to give an estimate of total effort. The units in which the total effort is then given will be the units in which the partial effort was given. Similarly the total catch could be calculated if the total effort of the fleet and the catch per effort of part of the fleet were known. The reliability of these estimates of course depends heavily on how closely the catch per effort of the whole fleet resembles that of the known part. In cases where the fleet as a whole is very heterogeneous but there is a sizeable group of reliable and non-varying vessels, the method of deriving total effort from the catch per effort of this group will be the best procedure to follow.
Length frequency distributions are normally the first step in looking at the detailed structure of the fish population, in particular for establishing its age composition and growth and mortality rates. This length frequency distribution of the landings is estimated by sampling, since it is generally impossible and unnecessary to measure every fish landed.
We noted in Section 2.2.1 that the biologist is often concerned with "gross catch" rather than "landings" or "nominal catch". Estimates of relative length frequency from landings will be biased when discarding is taking place, because it is normally the smallest fish which are discarded to comply with minimum size regulation or marketing requirements, The relative length frequency in the "gross catch" will not be the same as the length frequency in the fishing area because of the behaviour of the fish and the selectivity of the gear. In order to allow for this the mesh size and selection characteristics must be known for nets, and the pattern of avoidance for other types of gear. Finally the population in the area being fished may differ from populations in other arias in its size or age composition. In this case special surveys have to be carried out if data on the whole population are required, lost uses require only length frequency distributions of the "landings" or "gross catches", with some information on the mesh sizes in general use.
Length frequency distributions are mainly important as the first step in obtaining age frequency distribution, i.e., the numbers and sizes of fish of different ages in the "landings" or in the "gross catch". A series of age frequency distributions for a number of years forms the basis of most analytical assessment models. They are used to establish the growth of the various species of fish, the age structure of the population, the age at which young fish become liable to capture and how quickly the fish die off due to fishing and natural causes.
This kind of information, collected over a long period., is vital to an understanding of fish stocks, including the competitive relationships between different species, the relationship between the size of the adult stock of a species and the number of young produced (year-class strength) and the influence of climatic or other factors on year-class strength. These are the most important long term biological problems in rational exploitation of the sea, but the age and size characteristics are also needed by fisheries managers in order to take decisions on mesh size regulation and the closure of a fishery at certain times of year or in particular areas. Precise and effective regulatory measures of this kind can only he taken if sufficiently detailed routine statistics are available to evaluate their effect.
The other type of regulation being applied with increasing frequency is the annual catch quota and this too needs information on year-class strength, particularly of the in coming (recruiting) year classes. In a fishery which depends heavily on the recruitment of a fast growing species the catch for the next year can only be predicted accurately if the strength of this recruitment can be estimated. Up to the present time this has been done by special surveys, if at all, but if catch quotas are to become a regular management tool then estimates of the strength of incoming year classes will have to become a regular statistic. This type of information is obviously also of great interest to fishermen, particularly if presented in a fairly detailed and practical way (e.g., Figure 9 from Fishing Prospects, 1972 - 1973 (Great Britain 1972)). The methods of obtaining routine length and age frequency distributions of the landings are dealt with in Section 3. In situations where the fish cannot be aged it is often still possible to predict future catches from the size frequency distribution by using relative growth and mortality rates.
Figure 9: Catch-rates of Irish Sea cod 1968-72, and predicted catch-rate for 1973 (Great Britain, 1972).
We have already noted that the study of the fishing industry in its broadest sense requires data on many general aspects of the economy of a co as a whole, such as employment, income, population, investment etc. A discussion of all these requirements, important as they are to studies of the fishing industry, is obviously beyond the scope of this work. The collection of each statistics is usually part of the work of comprehensive national statistical systems such as industrial censuses, censuses of commercial establishments, imports and exports etc. To a large extent such system will cover the activities of the secondary and tertiary phases of the fishing industry: processing, marketing and distribution. The national fishery statistical system may have some responsibility to advise on appropriate and adequate coverage, but its main responsibility will be for statistics of the primary phase. The fisheries department, although not responsible for the collection of general industrial and social statistics, should be aware of their extent and of their shortcomings. For example in preparing national statistics of labour utilization and in come levels it is common for fisheries to be included with agriculture and forestry in the final publication. The value of such genera) statistics may be very low in a detailed economic study of the fishing industry, and the fisheries department should have access to any more detailed, intermediate statistics which are collected in preparing the overall figures, and should ensure that they are not destroyed once the overall statistic is tabulated., Cooperation between different users is essential to avoid duplication or wastage. Another common example 9 this is that statistics of numbers of vessels and characteristics will often be kept, for registration purposes, by the ministry or bureau responsible for all shipping.
Many of the statistics for economic and industrial studies are covered by the collect ion of data for stock assessment purposes. The differences lie in the amount of detail and importance of different statistics. For example, for many economic studies it is unnecessary to know the value of the landings of each different species, and a classification into "high value" and "low value" species may be sufficient. It is difficult to specify the amount of detail and the level of priority which should be given to different statistics without knowing the objectives for which they are needed. A few comments on the relative importance of different statistics within a category are given below but the choice of what to collect must be made in the light of each situation. It is probably a good idea to aim for some coverage in each category in order to follow general long-term trends.
The statistics included under the heading "Structure" are more or less unvarying and can therefore be collected once and updated as changes occur or on a regular basis. For example the number of vessels at a port or site could be counted once and checked annually, or it could be the duty of the harbour master, if one exists, to report any changes. The operational statistics need continuous monitoring arid will therefore be more expensive to collect. Opportunities for combining the collection of operational statistics needed by different users should not be overlooked. (e.g., effort statistics for biologists and cost of labour for economists can often be collected. at the same time).
It should be fairly obvious that some of the structural statistics listed here are needed before any others can be collected at all, because they provide the frame within which the others are collected:
1. The location and number of fishing Bites or ports; their size (number and frequency of landing which they can handle) and facilities (ice, processing plants, repair facilities, transport).
2. The number and characteristics of fishing vessels at each port or site. The characteristics may be adequately described, by giving the names of different classes of vessel, e.g., four different kinds of traditional canoe, or a complete breakdown of characteristics may be needed (see Table 4 for example). The number of vessels at each port or site must sum to give the total number of vessels engaged in fishing, i.e., it must take movements into account and not count the same vessel twice.
3. The value (purchase price and current value) end year of manufacture or purchase of all equipment used.
4. The number of fishermen. They can be classified by age and status. Three kinds of statue classification may be used. to show (a) time spent fishing, e.g., full time, part time, occasional, unemployed; (b) employment status, e.g., skipper, mate, deck hand, learner; (c) ownership status, e.g., owner, skipper/owner, shareholder fisherman, employee.
There are also other workers within the primary phase of the industry, who land the fish and do some of the grading and processing before it is sold. They can be grouped generally as "land based" or classified in ways similar to those used above.
Sport fishermen are excluded from international compilations of manpower s but their numbers are often recorded at the national level in order to measure the amenity value of particular resources. The classifications used in completing FAO forms on statistics of fishermen are given by FAO (1966).
The fishing inputs (e.g.; hours worked, supplies, vessel time) discussed in Section 2.2.2 are of interest from two points of view in studies of the industry:
1. measured in terms of their cost, they can be used to assess economic performance
2. the physical and engineering aspects of the fishing operation are measured for assessment of technical performance.
The two approaches may be integrated in order, for example, to evaluate the size and type of fishing vessel which will be most profitable in a given fishery situation (an example is given by Engvall and Engstrom, 1974).
Operational costs include:
(a) repair and maintenance
(b) supplies of fuel, ice, bait, lubricants and food at cost to the vessel
(c) labour costs and hours worked, which can be broken down according to any or all of the classifications previously given.
The frequency with which these statistics need to be collected varies. Repair and maintenance figures may be given annually, supplies and labour more often, in order, for example, to show seasonal variations in employment. The largest cost items are usually fuel and labour.
From both the economic and the technical points of view we need to know not only the cost or the amount of energy used, but the time over which it was used, e.g., the daily cost or rate of work. Since both cost and rate of work depend greatly on what the vessel is doing one needs statistics of time spent fishing, searching, steaming, unloading, in dock, being repaired and waiting for bad weather. Much of this information will also be needed
Table 4 : Characteristics of fishing vessels (FAO, Department of Fisheries, 1973)
1.0 VESSEL CHARACTERISTICS
The list of vessel characteristics to be retained in the register is given as follows:
(b) Registration number
(c) Base port
(d) Name of vessel
(e) Name of owner
(f) Type of vessel
1.2 Vessel characteristics
(a) Overall length (m)
(b) Overall breadth (m)
(c) Depth (m)
(d) Gross registered tonnage
(e) Year built
(f) Country in which built
(g) Hull conversion - year (with explanation on type of conversion)
(h) Building material for hull
(i) Propulsive engine (steam-piston, steam-turbine, motor (combustion), electric coupling (diesel electric), others)
(j) Fuel (coal, oil, petrol/gasoline, others)
(k) Capacity fuel bankers (m3)
(l) H.P. (total all engines)
(m) HP. for population
(n) Type of propeller(s)
(o) Propeller, revolutions per minute (free running)
(p) speed on trials (knots)
(q) Changes in propulsive engine
2.0 OPERATIONAL CHARACTERISTICS
2.1 Types of fishing gear
The fishing gears are recorded according to the classification shown in Part B of this Fisheries Circular.
The aim is to indicate the gear or gears for which the boat was built and not all gear. , could be used by the boat with or without transformations. The statistical record of fishing vessels cannot replace the operational reports on the catches taken by the vessels.
The actual gear utilised should be mentioned in the operational statistics; the fishing vessel statistics should only deal with the permanent characteristics of the vessel.
2.2 Personnel: (crew number)
2.3 Salting facilities :(yes or no)
2.4 Hold capacity (in cu.m.):
(a) Fish hold capacity
1. without insulation
2. insulated only
3. mechanically refrigerated.
4. frozen (only for frozen fish)
5. frozen (equipped to hold. also wet fish)
(b) Fish tanks
1. used dry or as water tank
2. water tank (refrigerated sea water or tanks for live fish, shellfish or bait)
1. liver oil
2. fish oil
3. fish meal
3.0 SPECIAL EQUIPMENT
3.1 Processing facilities
(a) Freezing (t. per 24 hours)
(b) Filleting, machine (number of machines)
(c) Filleting, hand (using specialized installation)
(d) Canning (t. of raw material per 24 hours)
(e) Meal (t. of raw material per 24 hours)
(f) Oil (t. of raw material per 24 hours)
(g) Others (t. of raw material per 24 hours)
3.2 Electronic equipment
(b) Radio V.H.F.
(c) Radar (number of instruments)
(d) Echo-sounder, vertical (number of instruments)
(e) Echo-sounder, horizontal or ranging (number of instruments)
(f) Decca navigator
(i) Direction finder
(j) Facsimile receiver
(l) Warp tension indicator
3.3 Other equipment
(a) Powered block
(b) Side thrust propeller(s)
(c) Fish pump
(d) Powered net drum
(e) Fishing with lights
(f) Electric fishing gear
in biological studies of fishing effort and unless there is a special requirement for a particular technical or economic study the biological requirement should be sufficient on a routine basis. The more detailed time statistics which may be needed for economic or technical studies can usually be collected ad hoc rather than routinely.
The output from the primary phase is fish and landed fish products of various kinds which are converted to figures of "nominal catch". They generate revenues to the fisherman, i.e., value at first sale, price. In order to measure technical or economic efficiency by means of input/output ratios these outputs must be classified into the same groups as the inputs (e.g., by vessel/gear type, port and time period), but in addition the total catch is usually classified first by species. The data on output available from collection programmes for biological work will usually be sufficiently detailed for economic studies, with the addition of price or value statistics, but three further cross-classifications may be needed:
1. Type of commodity landed
2. Disposition channel (see Table 5)
3. Quality-for marketing studies and price control
Input/output analysis of the later phases is aimed principally at establishing the value added at each stage and at charting connections with other industries and measuring capacities and flows. The inputs again are:
1. Capital - original cost, age, number, type and capacity of equipment and plant, and whether, for example, trucks can be used elsewhere in off-seasons.
2. Labour - hours worked and wages per month for different worker groups
3. Supplies -fuel, power, packing material, purchase of fish.
The purchase price in the secondary phase should equal the landing value, since this is the same transact ion, and the volume and value of sales from each phase should equal the volume and value of purchases later on. Any discrepancies should, be explained. The out put from the secondary and tertiary phases is measured by the quantity and value of the product. The stock holding capacity in each phase should be recorded.