3. THE MEASUREMENT OF FISH AND SHELLFISH

Measuring the linear dimensions of whole or parts of fish is probably the most widely used technique in fisheries biology studies. Of the many observations that can be made, the one that is most frequently chosen is that of total length because it is quick and easy to measure. Total length is usually closely related to many of the other factors such as weight, age and maturity so that any of these may be easily determined, from the length data used, with a small key sample relating length with the factor concerned. Length data collected from a developing fishery are particularly important; other biological studies can be made later, but measurement of sizes in the early days can be done only at that time and therefore should be given a high priority when a programme of investigation is being planned.

3.1 The Measurement of Fish

3.1.1 Choice of length dimension to measure

The choice of which length dimension to measure is arbitrary. Most fisheries biologists measure either fork length or total length but neither method has a clear advantage over the other sufficient to recommend its acceptance as a universal standard. However it is highly desirable that the same unit should be used by all those measuring fish from the same stock.

Damage to the tip of the tail can make it difficult to measure total length, but it can be as difficult to measure fork length when the tail is frayed and the fork is not well defined.

In most cases it is easiest to measure total length but for some large species with a rigid, deeply forked caudal fin, fork length is preferred. There is general agreement among tuna workers to measure the fork length. (Report of the FAO Expert Panel for the Facilitation of Tuna Research, Working Party on Tuna Length Measurements and Tabulation). The West African institutes concerned with the international Sardinella investigations in the East Central Atlantic have also adopted fork length as the standard dimension. On the other hand, the International Council for the Exploration of the Sea (ICES) and the International Commission for the North West Atlantic Fisheries (ICNAF) at their Sampling Meeting held in 1965 recommended that the length dimension to be measured for all species (except tunas and salmonids), for which length composition data are reported to the two commissions should be total length. This was defined as the maximum length as measured by bringing the longest lobe of the caudal fin into the mid-line of the fish (maximum total length). This standard covers most of the North Atlantic fisheries. This standard has since been accepted by most of the ICES member countries for most of the species, although it is still not universally in use. The 1968 meeting of the Regional Fisheries Advisory Commission for the Southwest Atlantic (CARPAS) approved the same standard for use in the Southwest Atlantic, and the General Fisheries Council for the Mediterranean (GFCM) (1963) adopted total length dimension for sardine. There thus seems to be already substantial agreement that fork length should be measured for tuna and mackerel, and maximum total length for nearly all other species.

Standard length is rarely used in fisheries studies, except by systematists. It is difficult to record accurately while working at speed and it has little to recommend its use in the field. Other standardized measurements in use by systematists are given in Appendix 1.

The important point to remember is to note whichever measurement you use and to state this in every publication, unless you are publishing in the statistical reports of a Commission which has an accepted standard. Failure to do so will inevitably result in confusion later. For example, one fisheries worker may measure fork length and a later worker, studying the same species, total length. The latter may assume that the first worker also used total length, if he did not say so, and draw false conclusions about changes in growth rates and length-weight relationships.

Whichever length measurement is chosen the metric system is used for recording it.

3.1.2 Methods of measuring

Measurements are made with special measuring boards, tapes and calipers and they are usually, but not always, taken along straight lines. Overall length measurements are usually made with the fish lying on its right side, snout to the left, on a measuring board consisting essentially of a wooden or metal base carrying a centre scale and having a headpiece (nose block) against which the snout is gently pressed. The mouth is closed, the fish body and tail are straightened along the mid-line and the reading taken from the scale. The fish should be measured while it is fresh and wet, that is, as near to the relaxed live condition as possible; fish will shrink rapidly on drying. Normally, when the fish are measured on fishing craft or in markets at the time of unloading, this problem will not arise, but if samples are taken away for later observation, a means of conversion of measurements to fresh, wet condition may have to be employed. Fish in rigor mortis (stiffness after death) should be flexed gently before they are measured. In certain circumstances, for example, during tagging experiments, fish may be measured alive. Then it may be necessary to use narcotics to relax the fish during handling. A solution of Sandoz MS 222 is widely used for this purpose.

Rays and other dorso-ventrally flattened fish may be measured while lying straight on their ventral surfaces. Disc width rather than overall length is sometimes used as linear dimension of rays. Large, fat fish present difficulties. Their lengths may be measured with calipers or from point to point along the body surface with a tape.

Commonly, measuring is done by observers working in pairs, one of whom places the fish on the board and reads the length; the other records the measurements called out by the first. A common source of error is indistinct calling by the measurer, causing the recorder to write down the wrong length. Usually fish are measured under adverse working conditions, either in fish markets or at sea, where there is frequently a lot of background noise. So, the first rule of measuring is call out the length loudly and distinctly.

Various techniques have been developed to allow one man to measure fish on his own. The most practical of these are the one-man measuring board and the use of a portable tape recorder. The one-man board is described in Appendix 2. It has superseded similar devices where lengths were recorded by pricking holes in strips of celluloid or aluminium foil fixed to the board. The use of a reliable portable tape recorder can be extremely useful at times but transcribing the information later can be time-consuming, and unless it offers any special advantage it is rarely either more efficient or simpler to use than the one-man board.

A measuring board consists essentially of a flat platform which has a measuring scale on it and on which the fish is laid. At the zero end of the scale there is a head-piece or stop, against which the head of the fish is gently pushed. It usually has a handle to make it easy to carry. It must be both strongly constructed and retain its accuracy after several years of use in adverse conditions (soaked in sea-water) and of being mishandled (dropped). The scale must be clearly marked and easy to read. If ink is used to mark the scale it must still be clearly visible after at least one year, preferably longer. The scale must correspond to the measurements being recorded. It is impossible to measure fish to the nearest centimetre below on a board marked 2 cm intervals. If the scale has too many divisions either mistakes will be made or time wasted trying to record to the nearest division.

For fish measured in centimetre units, a board 1 m long is usually sufficient, although it may be useful to have an extension piece 30 cm long, which can be clipped or hinged on, if large fish regularly occur in the samples. For fish measured in half centimetre units a board 50 cm long is usually sufficient. Do not use a board that is more than 10 cm larger than the largest fish regularly measured.

3.1.3 Measuring damaged and processed fish

Suppose that total length is the basic dimension chosen. If an occasional fish in a sample has its caudal fin damaged, it should not be discarded but its total length should be estimated by comparison with another fish of about the same size. This will avoid possible bias arising if damaged fish tend to be either larger or smaller than the average.

If whole samples or large proportions of them are damaged or processed, for example, if the heads have been removed or the fish have been dried, another convenient dimension must be measured. The Peruvian anchoveta are a typical example of this problem; the largest fish have the least damage; the smallest fish are often so badly damaged that it is impossible to measure their total lengths. Head lengths are measured instead. A method of arriving directly at total length from the measurement of another dimension is possible by the use of a specially constructed measuring board described in Appendix 3.

If some other dimension, such as head length, is measured for all fish then it is not essential to convert these measurements to total length. More often these measurements have to be combined with others of total length (as for the Peruvian anchoveta) and a conversion key must be constructed so that the total lengths of fish whose other measurements, the head for example, have been recorded can be combined with others whose total lengths have been noted. Conversion keys must be applied only to the stock for which they were determined and should be checked periodically to sea whether the key has changed. If it has changed the key must be recalculated unless some additional factors are found which allow the original to be used. As an example of possible change, the length of the head of a fast-growing fish forms a larger proportion of its total length than that of a slow-growing fish of the same size. The growth-rate is likely to increase as a stock is exploited and the abundance of the fish reduced. The original conversion key produced for the unexploited stock becomes invalid as the stock is exploited.

3.2 The Measurement of Shellfish

3.2.1 Molluscs

Bivalves (Fig. 3.1)

(i)  Length of shell. The greatest measurement is in an anteroposterior direction; this is usually approximately parallel with the axis of the hinge.

(ii)  Width of the shell. The greatest measurement is in a dorsoventral direction; this is usually approximately at right angles to the axis of the hinge and approximately at right angles to the length measurement.

(iii)  Depth of shell. The greatest measurement is at right angles to the plane of the above two measurements.

The measurements are made with either calipers or measuring boards. (Loosanoff, V.L. and Nomejko, C.A., 1949).

Fig. 3.1 Measurements of shellfish

Gastropods

The standard measurement is the maximum measurement from the tip of the whorl to the tip of the shell. (Warren, P.J., 1958).

3.2.2 Crustacea

Lobsters, crawfish, shrimps, prawns. The standard measurement is the minimum length of the carapace from the inside of the eye socket to the posterior margin of the carapace. In some cases (lobsters) the measurement is parallel to the mid-line and in others (some shrimps) it is from the eye socket to the centre of the dorsal margin of the carapace. (Cole H.A. and Mistakidis, M.N., 1953).

Crabs. The standard measurement is the maximum distance across the width of the carapace, including the spines if present, at right angles to the medium line.

All measurements are made with calipers.

3.3 Form for Recording Measurements

An example of the type of form used for recording measurement of many specimens of the same species is shown in Appendix 4. The main feature of this form is that only the units (0-9) are marked. The technique of using it is to establish the size range in the sample by selecting the largest and smallest fish visible. Say these fish are 55 cm and 25 cm long. A ‘1’ is then placed before the first zero to make it read ‘10’ and subsequent zeros made to read 20, 30, 40, 50 and 60. It should then be possible to record all fish within the sample. If a fish smaller than 10 cm or longer than 69 cm is found its length has to be noted separately. A form with 0.5 cm interval would be needed for recording fish to this length interval.

Each fish measured should be recorded in a square against the appropriate length, building up a series of windows of blocks in each square thus each representing five fish. This is a foolproof system, quick and easy to use and less liable to error than other methods, e.g. of ; with either of these methods it is possible to make simple mistakes either by inserting an extra vertical stroke or missing a single dot when totalling up the measurements.

The other side of the form is used to record details of the vessel, fishing gear, date and place of landing, area fished and quantity of all categories of the species landed.

3.4 Units of Measurement and Grouping of Data

The metric system should be used for all scientific measurements relating to the assessment of fish stocks. Choice of the class interval for recording length measurements is a matter of judgement but it is usually made in 1 cm units for species which grow larger than 30 cm and in 0.5 cm units for species which do not reach 30 cm. For very small species and for larval fish it may be necessary to work to 1 mm intervals but this is exceptional.

The method of reading length to the unit below has been agreed by the ICES/ICNAF Sampling Meeting 1965, the FAO Working Party on Tuna Length Measurements and Tabulation 1967, CARPAS 1968, the West African sardinella investigations 1967, and the GFCM sardine workers 1963. According to this method all lengths between 26.0. and 26.99 cm should, when recording in centimetres, be noted as 26 cm. Similarly when measuring in 0.5 cm units fish between 26.0 and 26.49 cm would be recorded as 26 cm and fish between 26.5 cm and 26.99 cm as 26.5 cm. This method has now become fairly universally agreed, at least in the Atlantic regions. It is far more satisfactory than measuring to the nearest unit, in which case 26 cm would mean all fish between 25.50 cm and 26.49 cm when measuring to the nearest centimetre and all fish between 25.75 cm and 26.24 cm when measuring to the nearest 0.5 cm.

It is often advantageous to combine length data into large groupings. The appropriate grouping interval for a species depends on the purpose for which the data are to be used and on the acceptable limits of accuracy of the estimates derived from them. Too fine grouping greatly increases the space needed in publication, the work of tabulating the data and the difficulties of calculations, without adding significantly to the accuracy of the information provided. A grouping such that the available length data fall into a range of about 20 units is sufficiently fine for most purposes.

In order to ensure that data can be pooled even when practices differ, it is suggested that the basic system of grouping agreed upon by the FAO Working Party on Tuna Length Measurements and Tabulation is adopted for general use. In this system, regardless of which unit of length interval is used, the groups always start at zero, and the only intervals used are 0.5, 1, 2 and 4 and 8 cm. Thus, the data given in the smaller units can always be combined into the larger groups if desirable for purposes of pooling with data given in other units, as is demonstrated in the following scheme:

Explanation overleaf

^*m = mid-point of the length unit; fr = imaginary size distribution used as example

This scheme also shows the figures of the mid-points of each grouping. These mid-points have to be taken into account in calculations such as that of average length, etc. It should be made clear in the tabulation of size distributions whether the sizes given refer to the unit below or to the mid-point of the grouping. A convenient way, which leaves no ambiguity, is shown in the following example:

26 - means greater than or equal to 26 but less than 27, i.e. incorporating all intervals from 26.00–26.99, etc.

This scheme demonstrates the advantage of reading to the centimetre below compared with reading to the nearest centimetre. With the latter method 26 cm would mean all fish of lengths 25.50 cm to 26.49 cm.

One practical point; when calculating mean lengths of fish measured to the nearest centimetre below it is necessary to add 0.5 cm to the mean length calculated. The reason for this is that some fish will be only 26.0 but others will be as long as 26.9 cm. The mid-point will be 26.5 cm.

3.5 Reporting of Length-Composition Data

Reporting of age- and length- composition data for inclusion in Biological Sampling Yearbooks and in general for assessment purposes should be done in a standardized form which includes all essential information. Some general criteria have been developed and are in use in other international organizations.

There are two sources of length-composition data; those obtained from:

(a) comprehensive market sampling programmes, and

(b) occasional sampling of commercial landings and research vessel catches.

For assessment purposes the first of these is usually most important, but data from both of these sources are valuable and should be reported, clearly specifying the category to which the data belong. If possible, data from source (a) should be expressed as the length-composition of the total landings of all that species for the particular gear, area and time interval. They should also be accompanied by the following additional information:

i) number and size category of vessels sampled;

ii) number or weight of fish measured;

iii) total number or weight of fish landed by the fishery for the particular gear, area and time period.

An example of the reporting form used by ICES is shown in Appendix 5.

The data from source (b) should be submitted as per mille frequency distributions, and should be accompanied by the following additional information:

i) source (e.g. research vessel, commercial vessel, etc.);

ii) gear used (including mesh or hook size);

iii) number or weight of fish measured;

iv) number or weight of fish caught or landed.

3.6 Weighing

Weights are as good as length measurements but they are not used routinely because they are time-consuming to collect and liable to large inaccuracies. However, it is essential to establish length-weight relationships for all species because at some stage in stock assess ments lengths of fish have to be converted to weights of fish. It is useful to have seasonal length-weight relationships for both whole and gutted fish; this is particularly useful for whole fish whose weight may be much greater during the spawning season than during the non-spawning season.

To convert ‘landings’ to ‘nominal catch’ it is essential to establish relationships between weights of fish as landed and whole weight for all ways in which the fish is landed, such as gutted, headed, filleted.

Weights of various parts of the fish may be required for biological studies; stomachs for the determination of food eaten, gonads for the estimation of egg numbers and livers for liver-condition assessment.

The only reliable transportable instruments are steelyards and spring balances. The latter are sensitive to temperature changes and many other factors, but small pocket-size spring balances in various ranges (e.g. 0.1 kg and 0–10 kg) are useful for rough measurements and should be carried on all field trips. Simple transportable laboratory balances are also useful but the weights with which they are used make them slow in use. Shop balances can also be used in fish markets, if a stable base can be provided and care is taken of them in transport. Excess water must continually be removed from the pans of these balances or the weights will become increasingly inaccurate during a recording period.

It is usually so difficult to weigh fish at sea that samples are often preserved by either deep freezing or placing in formalin for later weighing ashore. To avoid losses in a deep freeze fish must be placed in tightly sealed polythene bags. For fish preserved in formalin fresh-preserved weight conversion factors must be established.

3.7 References

Cole, J.A. and M.N. Mistakidis, 1953 Advice for the quick and accurate measurement of carapace length in prawns and shrimps. J.Cons.Perm.Int.Explor.Mer, 19(1):77–9

Loosanoff, V.L. and C.A. Nomejko, 1949 Growth of oysters, O. virginica during different months. Biol.Bull.Mar.Biol.Lab., Woods Hole, 97(1):82–94

Warren, P.J., 1958 Advice for rapid single-handed measurement of shellfish. J.Cons.Perm. Int.Explor.Mer, 23(3):440–2

Appendix 3.1
Definition of Linear Measurements

Overall lengths may be measured from the snout (U, the position of the maxillary symphysis) or from tip of the lower jaw (L, the mandibular symphysis). Measurements from L are taken with the mouth closed. If the lower jaw projects much beyond the upper jaw, measurement from S may necessitate provision of a special “stepped” nose piece on the measuring board.

There are three main overall length measures in common use. These are roughly illustrated in Figure 1. Standard length, the standard dimension of taxonomy, is properly measured from U to the tip of the hypural bone (urostyle). In practice it may be measured to some external feature more or less corresponding with the latter point; this would vary from one species to another but examples are the last scale, or point of silvering, to edge of akin pigment (determined by scraping away the posterior scales), to tip of fleshy peduncle or to the keel. Fork length is measured from U or L to the cartilaginous tip of the shortest, or median, caudal fin ray. It may be difficult to measure if the tail is split. Total length is measured from U or L to the tips of the longest caudal fin rays, in several possible ways. Measurement may be taken to the tip of the dorsal lobe, of the ventral lobe, or of the longer of the two, or some average of them both (such as to the midpoint of a line joining the two tips or to the point where such a line crosses the median longitudinal axis).

The tail fin may be in the extended position to give normal (total) length or the tips of one or both caudal lobes may be drawn to the longitudinal axis extreme (total) length. Sometimes total length measurements are made with the caudal lobes partially drawn together so that their outer edges are parallel to each other and to the axis. It is therefore clear that precise instructions must be given as to the dimension to be measured and the mode of measuring. In particular, if a normal length is the chosen dimension, it is necessary to standardize the procedure of laying the fish on the board. A common method is to place the head of the fish against the nose piece with the right hand, hold the fish in position with the left hand, and use the right hand to straighten the body of the fish and extend its tail with a single stroking movement.

There follow useful sets of notations and definitions for positions and linear measurement, which are applicable to many types of fish; they are illustrated with reference to a scombroid.

Positions are defined as on the left side of the fish unless stated to the contrary. When side-to-side comparisons are being made, or if necessary for other reasons, distinguish by prefixing “r” or “g” (right or greater) to the notation (or term). This rule may be especially applicable to the dimensions Ph and Vh.

Lower-case letters are used by themselves in the notation where the dimension is not defined with respect to two fixed points, e.g., q, b, g. Otherwise d stands for “diameter”, h for “depth”, and g for “greatest”.

Overall length measurements are always made between perpendiculars along the median longitudinal body axis from L or U with mouth closed. Measurement is understood to be from L unless otherwise specified, thus, e.g., UX Upper (or “maxillary”) dorsal extreme length.

Fig. 1 Body measurements of fish

Longitudinal measurements other than overall length are also often made between perpendiculars, using a measuring board with, for example, a sliding cursor. When they are made radially from point U, using calipers as is recommended, the symbol for the dimension should be in parentheses, and the name followed by the qualifying term “direct”. Point-to-point measurements are sometimes made on big fish, e.g., tunas, by tapes. They would be indicated by the word “surface”; they are not generally recommended. The term “radial” is not recommended to be applied to point-to-point measurements as it invites confusion with finray counts.

All measurements listed from LX to LM and also their “upper” equivalents are grouped under the general name “total length”, LT. LM has also been called “bilobular length” and “total auxiliary length”. It is rather difficult to measure.

The word “extreme” is used in LX, LX' and LX'' instead of “maximum” to avoid confusion with the asymptotic size referred to the growth studies. “Greatest” is a possible alternative to “extreme” but invites confusion with “greater” for which a suitable alternative cannot be found.

LF and LF' have also been called “median length” or “midcaudal length”. The term “standard length” has on occasion been applied to the distance LB as well as to the LS series; this usage is not recommended. It is not known that the ventral body length, LB', has ever been used in practise.

“Depth” is used in terms from Oh to q instead of the alternative “height” to avoid possible confusion with fin heights as measured by length of rays. The ambiguous term “width” is not recommended. Again “width” is not recommended as an alternative to “breadth” in terms for PP and b but “thickness” would be suitable.

Pectoral and ventral fins, Ph and Vh, are usually measured in their folded positions opposed to body side (to keep rays straight) from foremost visible point of insertion to the distal tip of the membranous edge.

In measurements of eyes the limits of Id, Ih and Ig are the boundary of the iris with the black tissue. Ig and Eg are not necessarily equivalent to either Id or Ih, Ed or Eh respectively.

The method of measuring g, in particular the permissible degree of constriction of the belly, remains to be defined precisely. Other girth measurements might be more consistent and perhaps related to conditions relevant to selectivity problems, e.g., if taken level with Y or D1. An appropriate notation could be gY, gD1, etc.

Spread caudal distance NN' has been used for tunas.

In various types of fishes certain of the dimensions listed may be virtually indistinguishable. Thus in the mackerel Rastrelliger LM = LM', LS = LS', UJ = UO', UY = UJ', and so on.

Appendix 3.2

INDO-PACIFIC FISHERIES COUNCIL

Occasional Paper 60/3

“ONE-MAN” MEASURING BOARD

by

T. Williams, Fisheries Laboratory, Lowestoft

During the FAO International Training Centre on the Methodology and Techniques of Research on Mackerel held at Bangkok in 1958, the possibility of producing a “one-man” measuring board was discussed. It was decided that the board needed to be simple to use, inexpensive to produce, and one that would enable a research worker to measure and record, single-handed, large numbers of fish under field conditions.

The Fisheries Laboratory at Lowestoft, England, made and tested a board according to the principles laid down at the Training Centre and found that it fulfilled these requirements.

Figure 1 is a drawing of the board showing the scale marked in centimetre units, and a line drawn down its median axis to help keep fish straight during measuring. The scale is drawn using an ordinary pen, and printers' ink Grade HJ. 150/A, on a white, matt surfaced, plastic material, which is attached to the wooden board with an impact adhesive (“Evostick” was the adhesive used).

The plastic surface is 0.02 inch thick and the material has the trade name “Astrafoil”. It is supplied with one side polished and one side matt, and it is necessary to roughen the polished side before attaching it to the wooden board.

Parana Pine, a knotless wood which is easily obtainable locally in England, was used to make the measuring boards, but a hard wood would be equally suitable. It was not found necessary to treat the wood of the boards that were used during the trials, but it may be found desirable to varnish it to prevent warping, should the boards be subject to continuous soaking, followed by storage under conditions of high temperature.

When measuring fish, each one is laid on the board with the head touching the upright headpiece. The fish is straightened so that it is parallel with the median line, and the length is recorded by a pencil mark in the centimetre space within which lies the extreme end of the tail or other point on the fish from which the measurement is to be taken.

The length distribution of the sample is built up by recording the lengths of the fish in this way. The marks, each of which represents a fish, are usually arranged in blocks of five, either or .

Two or more samples may be measured without the board being cleaned, provided there is sufficient space for recording, and the samples are clearly separated.

Figure 2 shows a series of lengths recorded on the board and a fish being measured. Sample 1 has been completed, and is separated from Sample 2 by a line drawn by the measurer.

The method of measuring described above results in the fish being measured to the centimetre below its actual length, e.g., a fish of 25.8 cm is recorded as 25 cm. If it is required to measure a fish to the nearest centimetre, e.g., a fish of 25.8 cm being recorded as 26 cm, then this can easily be achieved by screwing a 5 mm block to the upright headpiece, as illustrated in Figure 1.

A soft pencil is used for recording on the board and the marks can be removed with an ordinary pencil eraser. If, after a period of use, the board becomes difficult to clean thoroughly, xylol applied with a soft cloth and rubbed hard will remove all the pencil marks. This should not be done too frequently, as xylol removes a little of the ink each time it is applied.

“Astrafoil” supplied by D.E.P. Ltd., Frith Park, Walton-on-the-Gill, Tolworth, Surrey.

Printers Ink, Grade HJ. 150/A supplied by Hellerman Ltd., Crawley, Sussex.

Note by S.J. Holt (FAO)

Since this paper was written the measuring board described has been used with success by Mr. G. Tront aboard the FAIRTRY, a large British factory trawler. At FAO a soaking test has shown that “bubbles” between the plastic surface and the wooden base removed themselves when the board is immersed for a few hours; thus wetting improves the board, at least in this respect.

FAO has distributed 20 models of the board among IPFC member countries for field tests, the results of which it is hoped will be reported to the Council.

Author's Note

Later versions of the board have been made with resin-bonded marine plywood approximately 18 mm thick; this material is less liable to warp and has a longer life under field conditions.

Fig. 1

Fig. 2

Appendix 3.3

“A Method of Measuring Headless Fish” (Reprinted from J.Cons.Perm. Int.Explor.Mer, 31 (2): 279-83)

A common problem for the fishery scientist is to measure some attribute of a population. The study of catches of fish by commercial vessels for species, size, age, maturity or other characteristics often involves the organization of some sampling system. Of the many observations that can be made, the one that is most frequently chosen is that of total length which can be quickly and accurately measured. Total length is usually closely related to many of the other factors, e.g., weight, age, maturity, etc., so that any of the latter may often be most easily determined from the length data plus a small key sample relating length with the factor concerned.

The difficulties usually encountered when trying to obtain the length composition of a commercial catch are those of planning and organization, e.g., deciding the size of the sample to be measured, the frequency of sampling, and the stage at which the measuring should be done, e.g., whether at sea, immediately after landing, or later - after sorting and selling.

If it is decided to measure the fish ashore, an additional complication sometimes occurs i.e., the heads of some of the fish may be removed at sea before stowage, and part of the catch landed headless. The problem then is the difficulty of finding out what was the total length of all the fish.

This problem has recently been encountered in South Africa, where a team from the Lowestoft Laboratory is carrying out a survey of the stocks of hake (Merluccius capensis) in the area. It was found, however, that there are linear relationships between the total length (Lt) of the fish and the measurements Lp (the length from pectoral fin to end of tail) or La (the length from anus to end of tail). These relationships were therefore calculated as follows: a sample of over two hundred fish, covering the whole of the length range, was taken, and on each fish measurements were made of Lt, Lp and La. Mean values of Lp and La were plotted against total length to give two straight line relationships, (Lp=0.728 Lt+0.788; La=0.568 Lt+2.293) which are shown in Figure 1.

These relationships were then read from the graph (Figure 1) and used to construct a specially calibrated measuring board that immediately converts Lp or La into total length. For the Lp calibration the board is graduated in units of 7.28 mm, each unit being equivalent to 1 cm of total length; thus if Lp measures 35 cm this is equivalent to a total length of 47 cm and this is read off directly from the board (Figure 2). Other examples are given in Figure 3. The factor for La is 0.568, and for this conversion the board would be graduated in units of 5.68 mm. One board could be graduated to measure both Lp and La (Figure 4).

In this way the total lengths of large numbers of headless fish are measured quickly and easily, and a size composition of the catch by total length is directly recorded, without the need of any further conversion. It is obvious that the relationship Lp or La to Lt will vary between species, or possibly between similar species from different areas, and new values of the function of Lp and La to Lt should be established when any new stock is sampled.

When measuring headless fish a board without a headstock is necessary because the base of the pectoral fin is placed against the end of the board, the fin being held at right angles to the fish. However, a normal board with headstock can be used in the usual way for measuring total length of whole fish, whilst the end away from the headstock is calibrated to measure total length of headless fish by means of either Lp or La (Figure 5).

The most suitable type of measuring board for this work is the “one-man board” developed at the Fisheries Laboratory, Lowestoft; this board is surfaced with a matt white plastic (Astrafoil), is cheap to produce, quickly and easily calibrated, and has proved most satisfactory in use. It is fully described in the FAO Indo-Pacific Council Paper No. 60/3. The measuring boards used in South Africa were made of 1.9 cm resin-bonded marine plywood, which has proved less likely to warp than other types of wood.

The Lowestoft board can be used in the ordinary way with two men, one recording and one measuring, but its main advantage is that it can be used as a one-man board, the measurements being recorded directly on to the board by the measurer. A soft pencil is used for recording and will write on the plástic-surfaced board through water or fish slime. A pencil mark is made in the centimetre space within which the end of the tail lies, and the marks are usually built up in blocks of five as shown in Figure 6. When a sample has been complete it is a simple matter to transfer the numbers of fish, marked on the board at each length, to a log book. The board is then cleaned with an ordinary india rubber and the procedure repeated for subsequent samples.

This method is similar to the one described by Davenport and Harling (1965), but instead of transferring the recorded measurements from the plastic board to a log book, they used a new graduated plastic strip, pinned to the board, on each occasion.

Reference

Davenport, D. and W.R. Harling, Method of rapid measurement for large samples of fish 1965 J. Fish,Res.Bd Can., 22:1309-10

Figure 1. The relationships between L_T and L_A for South African hake. L_T = total length; L_P = length from pectoral fin to end of tail; L_A = length from anus to end of tail.

Figure 2. Measuring board graduated in units of 7.28 mm to give total length by measuring L_p.

Figure 3. Three examples of L_P being directly recorded as total length.

Figure 4. Board without headstock graduated to record total length from both L_P and L_A.

Figure 5. A. Measuring board graduated to measure both total length of whole fish and also total length (from L_P) of headless fish. B. Positions of the fish on the board.

Figure 6. The method of recording on a one-man measuring board.

Appendix 3.4

Example of the type of form used for recording measurements of many specimens of the same species

Appendix 3.5

Standard form used by ICES for reporting measurements of fish landings from regular market sampling programmes

1) 1955 and older