The determination of the sex ratio and of the sequence of changes in maturity stage during the year are of considerable importance in building a thorough knowledge of the general biology of an exploited stock. These form part of the basis of stock assessment. For some species it may be necessary to maintain routine programmes of sex ratio and maturity stage analyses. The male and female fish of some species, such as North Sea plaice, Tilapia and Sebastes, have such different rates of growth that they should be treated as separate stocks in stock assessment work. Mortality rates may also differ between sexes. Moreover, where the catch of a species contains a mixture of stocks, maturity data may provide the best guide to the relative proportions of the stocks in the catches and to changes in these proportions.
However, the determinations of sex and sexual maturity stages find their primary application in providing basic knowledge of the reproductive biology of a stock. The information derived from these analyses can be used in ascertaining the age and size at which fish attain sexual maturity, the time and place of spawning and the duration of the cycle from the beginning of the development of the ovary to the final release of eggs. Together with fecundity estimates this information can be used to calculate the size of a stock and its reproductive potential. The data have several practical uses. The age and size at sexual maturity may be important in assessing the optimum age of first capture of a species and the time and place of spawning can be used to plan fishing tactics because many species of fish are easiest to catch when they congregate to spawn. Conversely it may be considered advisable to limit fishing on an overexploited stock in which future recruitment is jeopardized by a low spawning stock.
The two major groups of fish, teleosts and elasmobranchs, differ so much in their reproductive biology that they will be dealt with separately. Most of the literature is concerned with the description of teleost fecundity and almost all of that is limited to description of species which lay all their eggs within a short period.
Determination of sex does not normally present any serious difficulties. In some species, e.g., salmon, characins and some cichlids, this can be done from external characters without opening the fish to expose the gonads. It is possible to sex plaice, and many other North Sea flatfish by holding them towards a light; the body cavity of the female is longer than that of the male (Fig. 5.1). Normally the body cavity has to be split open. Even after exposure of the gonads, differentiation between the sexes by gross examination may be difficult or impossible in small virgin individuals, as shown by Howard and Landa (1958) for Anchoveta and by Schaefer and Orange (1956) for skipjack and yellowfin tuna. In specimens which are beyond the immature virgin stage the distinction between sexes can normally be made easily by eye examination; ovaries are usually tubular, pink, and granular while the testes are flat, white and their ventral edges frequently have a wave-like outline. In other species, for example herring, the sexes of virgin fish can be distinguished from the colour of the gonads; the ovaries are red and the testes are white-grey/brown. The testes also have a more flattened knife-edged shape than the ovaries.
The term ‘maturity stages’ has a peculiar, but generally accepted meaning in fisheries biology. It is taken to mean the degree of ripeness of the ovaries and testes of a fish and not whether the fish has sexually matured or not. Thus the term ‘first maturity’ is used to Rastrelliger appears to meet most of the requirements and to be applicable, with minor modifications, to a wide range of partial spawners. This is given in Table 5.2.
Determination of maturity stages by visual examination using maturity keys lacks precision because it relies upon subjective judgement. It is adequate for many purposes but a more precise and more objective method is desirable in some causes. The most practical way of achieving this with the minimum cost and labour is to calculate a ‘gonad index’. This can be expressed as
Where w = weight for both gonads (g), and L = total length of the fish in millimetres. Since the weight of most fish is closely proportional to the cube of the length this gives an index which is approximately proportional to the relative weight of the gonads. For female fish it is also a relative measure of the ova diameter, independent of the length of the fish (Schaefer and Orange 1956).
It is not normally possible to do maturity sampling on commercial fish markets; often the fish are landed with the viscera and gonads removed. Also, fish markets do not normally have the facilities required for this type of work. It is therefore customary to do maturity staging on special samples either set aside ungutted for this purpose by the crews of commercial vessels or collected on research vessels. These samples may not be completely representative of the total catch because there is often a close relation between maturity stage and the length of fish. Therefore it is better to construct a maturity-length key and to use this to estimate the maturity stages in the population.
Exactly the same method of doing this is used as for constructing an age-length key (see section 4.4.2) except that the number of fish of each maturity stage replaces the number of fish per age group in each length stratum. The method of raising the maturity-length key to obtain the maturity stages in the total number of fish landed is then calculated in the same way as raising an age-length key to the age composition of the total landings.
The distribution of maturity stages in the commercial landings may not be representative of the distribution of maturity stages within the population in the sea. This can occur because fish in different maturity stages are not equally available or vulnerable to the fishery. For example, ripe herring are poorly represented in the catches of the Scottish driftnet fleet from the north-western North Sea. There are two reasons for this; ripe fish do not arise at night to the depth fished by a drift net to the same extent as fish in lower maturity stages, and the spawning grounds in this area are difficult to fish by drift net. Similarly, Schaefer and Orange (1956) have reported that yellow-fin tuna become unavailable to the American fishery in the Pacific on the approach of spawning, presumably because they then move either offshore or into deeper water to spawn. Such phenomena should be apparent in the results after sampling throughout a complete sexual cycle. They can probably be rectified only by appropriate research vessel sampling.
Knowledge of the fecundity of a species is an important factor in fish stock management. It is used to calculate the reproductive potential of a stock and the survival from egg to describe a fish which is spawning for the first time. For all other animals the term ‘maturity’ is used because an animal reaches maturity (the ability to reproduce) once. ‘First maturity’ implies more than one ‘maturity’. The inconsistency of the expression and the use of the term ‘maturity stage’ probably arose because the first fish for which ‘maturity stages’ were described had one, clearly marked, annual breeding cycle with a long interval in which the gonads returned almost to their virgin stage. However, it would be more logical to talk of ‘maturity’ and ‘spawning’ stages.
Routine assessment of maturity stages is normally done by assigning individuals to stages by characters which can be differentiated with the naked eye. A more refined distinction between stages can be made by histological examination but this is not a practical approach in routine sampling because it takes too long. The aim should be to examine a large number of fish at frequent intervals to get a representative picture of the stage of maturity of the population and the changes in this with time.
A large number of keys for maturity staging have been devised and described in the literature. These cover the minor differences between species and those within a single species, giving various degrees of refinement. Because gross naked eye staging inevitably means subjective judgement, too high a level of refinement is unjustified. A soale of not more than 8 stages is probably suitable for most species.
‘Total spawners’ are those species in which, after maturation of the gonads begins, all the eggs or sperm which are going to be spawned by the individual fish in a single breeding period develop synchronously. Their release takes place over a short period of a week or so and the breeding season is clearly defined. This is the commonest type, at least in species of northern latitudes.
Maturity staging of total spawners is usually simple because nearly all the developing eggs in the ovary are at the same stage and can fairly easily be allocated to that stage on visual criteria of size, colour, and texture, although these stages may have no very clear histological significance. A typical and fairly satisfactory key for classifying maturity stages in total spawners which is widely applicable in species of this type is that of Maier (1908) given in Table 5.1. This can be modified to suit the species under study. For example it is often difficult to separate males into 8 stages.
This scale adequately meets the basic requirements. It (a) distinguishes in low maturity stages between virgin fish and fish which have spawned previously, thus providing the possibility of fixing the mean age and range of ages at maturity; (b) clearly defines the stage both when release of eggs and milt is in progress and when it is completed, allowing accurate fixing of the beginning, peak, and end of the spawning period; (c) divides the intervening period into a reasonable number of stages, from which an approximate prediction of spawning time can be made and (d) is capable of rapid determination with the minimum of equipment, thus permitting large samples to be analyzed under field conditions.
‘Partial spawners’ are those in which spawning by individuals takes place over a protracted period and in which ripening eggs at very different stages of development can be found at any one time in the same ovary both before and during spawning. This situation is found, for example, in North Sea mackerel, in sprats and in a number of species, such as Rastrelliger and Chilean hake, in tropical and sub-tropical waters.
The construction of a maturity scale for partial spawners is more difficult because there is a range of development stages in an individual gonad at any one time and the differentiation achieved will inevitably be less precise. A key originally devised for recruitment, on both of which a judgement of the minimum adult stock necessary to maintain recruitment can be made. A knowledge of fecundity and the sex ratio of the adult stock are also needed to calculate the size of the stock from estimates of annual egg production which will be described in section 7. A third use of fecundity data is to discriminate between stocks in fisheries exploiting a mixture of two or more stocks with different fecundities (Baxter, 1963).
The problems of estimating fecundity depend upon several factors: the absolute number of eggs produced; whether the species is a total or partial spawner; on the degree of differentiation there is between the size of eggs which will be spawned in that season and any immature eggs present which will be carried over to the next spawning season.
As in maturity stage estimations the simplest case is that of total spawners. Fecundity can be estimated by removing the ovaries from females in Stages III to V of the scale given in Table 5.1. The fish from which they are taken should be measured and the necessary parts taken for age determination. The ovaries should then be preserved in modified Gilson's fluid (100 ml 60% alcohol, 800 ml water, 15 ml 80% nitric acid, 18 ml glacial acetic acid, 20 g mercuric chloride) with a label to identify them with the fish from which they were removed. The ovaries should be shaken periodically whilst in the Gilson's fluid to help loosen the ovarian tissue and to ensure rapid penetration of the preservative. After at least 48 hours in preservative the eggs can be completely liberated from the tissues by vigorous shaking.
The most accurate way of estimating the number of eggs in the ovaries is to count them all. This can be done with egg-counting machines but because the fecundity of most fish is so high this is impractical to do manually. Instead, it is necessary to estimate the number of eggs by sub-sampling. There are two basic methods of doing this, gravimetric and volumetric.
Gravimetric sampling as its name implies is based on weighing the eggs. After the eggs have been liberated from the ovarian tissues, they are thoroughly washed and spread on blotting paper to dry in air. The total number of eggs is then weighed and random samples of about 500 eggs are counted out and weighed. The total number of eggs in the ovaries is then obtained from the equation F = nG/g where F = fecundity, n = number of eggs in the subsample, G = total weight of the ovaries, g = weight of the subsample in the same units.
If Gilson's fluid is not available it is possible to estimate fecundity using this method by weighing both ovaries and then taking subsamples which are weighed. The eggs are teased out with a pair of needles and counted. At least three subsamples should be taken from each ovary, one from the anterior, one from the middle and one from the posterior. The method is tedious and less accurate than if it had been possible to use Gilson's fluid. Its great advantage is that it does not require large volumes of an expensive fluid and can be used in the field. It is particularly useful for fish with low fecundity and large eggs.
The volumetric method is very similar. After separation in Gilson's fluid the cleaned eggs are put in a measuring cylinder and made up to a known volume with water. Subsamples are then taken by shaking the container until all the eggs are evenly distributed through the water, a subsample of known volume withdrawn with a Stempel pipette, and the number of eggs in the subsample counted. The fecundity is then F = nV/v where n = number of eggs in the subsample, V = volume to which the total number of eggs is made up and v = volume of the subsample.
In practice, it is normally necessary to count more than one subsample from each fish to get a reliable estimate of the fecundity. Replicate counts of subsamples from the same ovary show that the distributions of the individual counts are of the Poisson type. Because the variance in a Poisson distribution is equal to the mean, the number of subsamples required to give any desired degree of accuracy can be determined from the number counted in the first subsample from the equation, % accuracy = 100/m where the mean, m, is taken as the count in the first subsample and n is the required number of subsamples.
This method is subject to considerable bias because it is very difficult to get all the eggs evenly distributed throughout the measuring cylinder. Unless great care is taken the density of the eggs is higher at the bottom of the cylinder than the top and in the middle of the cylinder than at the sides. The degree of bias is likely to be greatest if one person has to shake the cylinder and take the subsamples.
Total counts of egg numbers in an ovary can also be made using automatic counters either of the type described by Parrish, Baxter and Mowat (1960) (Fig. 5.2a) or the Decca Master-count type described by Boyar and Clifford (1967) (Fig. 5.2b). The advantage of using these machines is that sampling error, inherent in any subsampling technique is avoided but their disadvantage is their slowness.
Automatic fish egg counters, based on the coulter counter, are now being designed which enable the numbers of particles of several pre-set sizes to be counted. This will eliminate one of the drawbacks of present counters, which count all particles, large and small. In the new counters the number of particles corresponding to the size of the eggs only is taken as the fecundity.
To date, fecundity analysis has been largely confined to total spawners because it is so difficult to estimate the fecundity of partial spawners. In their early maturity stages, all the oocytes due to be spawned in one spawning cycle may not yet have been differentiated and, in later stages, some of the first eggs to develop may already have been spawned. To get an adequate estimate of annual fecundity in such species, it is necessary to obtain information on the number of spawnings per year, the number of eggs shed at each spawning and the relation between these factors and the size and age of the fish. Fischer and Balboutin (1970) have described useful techniques for sorting oocytes into size groups and subsampling them in such species while Macer (in press) has described a method for the horse-mackerel (Trachurus trachurus) based on histological examination of the ovaries over a complete spawning cycle. His method is very time-consuming and is not applicable to routine sampling programmes but it does allow a solution of the problem.
The sex of elasnobranchs can always be determined from external characters because male fish have a pair of mixopterygia (intromittent organs, claspers) which are visible from an early stage of development on the inside edge of the pelvic fins (Fig. 5.3). The females do not have mixopterygia.
The maturity of males can be easily and best defined from the state of development of the mixopterygia. These of immature fish are small and flaccid and do not reach the posterior edge of the pelvic fin (Fig. 5.3a). In maturing fish the mixopterygia are larger; they extend to the posterior edge of the pelvic fins and the internal structure is visible but soft and not ossified (Fig. 53b); in mature fish the mixopterygia extend well beyond the posterior edge of the pelvic fin, the internal structure is visible and is hard and ossified (Fig. 5c).
Maturity of females must be determined by internal examination. The reproductive system of females consists of ovaries (usually two but in some species one only is present), shell glands and oviduots (Fig. 5.4). In immature fish the ovary is barely discernible and it contains no eggs; the shell gland is also very small and the oviducts are thick-walled and white (Fig. 5.4a). In maturing fish white eggs are visible in the ovary but the remainder of the reproductive system is similar to that of immature fish (Fig. 5.4b). In mature fish the ovaries contain yellow eggs, except immediately after ovulation in viviparous species and at the end of the spawning season in oviparous species; the shell gland is enlarged and the ovidcuts distended and, in viviparous species, thin-walled, flaccid and often highly vascularized (Fig. 5.4d). In viviparous species maturity is also associated with changes in the size of the cloaca (Fig. 5.4c).
Elasmobranchs are either oviparous (egg laying) or ovoviviparous (the eggs are retained within the mother until the young are capable of free existence but no food is supplied by the mother to the young) or viviparous which is similar to ovoviparity except that there is some type of connection between the mother and the young by which they are fed.
Determination of the fecundity of the two live bearing groups depends upon knowing both the number of young and the length of the pregnancy cycle. The former can be determined by counting the developing, yellow eggs in the ovary or the number of young in the ‘uterus’ (the combined oviducal-cloacal system). In the latter instance care must be taken to ensure that the young have not been prematurely aborted during capture and in both cases it must be established that cannibalism does not take place in the uterus. (In some species the first young to hatch eats the remaining eggs). The duration of the pregnancy cycle is less easy to determine. The best way is to obtain samples over one year and to determine the average length of the young and the volume of their yolk sacs. Any inconsistent sudden increases in the growth of the young should be treated with suspicion. Pregnancy cycles of live-bearing elasmobranchs are often very long, two years for the spiny dogfish Squalus acanthias (Ford, 1922). Often the females in one stage of the cycle are not easy to catch and this can lead to false conclusions about the duration of the cycle (see Holden (1974) for a full discussion of these problems).
Oviparous species are partial spawners and pose the same problems as teleost partial spawners. The only study to date of this problem is that on the thornback ray (Raja clavata) by Holden (in press). He determined the occurrence of egg capsules in samples taken throughout the year and then equated the month in which the occurrence of egg capsules was at a maximum with the maximum rate of egg laying observed in aquaria (1 egg every 24 h). For other months the rate of egg laying was assumed to be proportional to the percentage occurrence of egg capsules. This, multiplied by the number of days in the month, gave the number of eggs paid in that month. For example, let the percentage occurrence in April (30 days) be 40% compared with 80% in June, the highest percentage occurrence of egg capsules observed. 80% is taken as equivalent to one egg laid in 24 h and therefore 40% is equivalent to 1 egg laid every 48 h which in 30 days would result in 15 eggs being laid.
In all species fecundity appears to be related to the length of the fish by an equation of the type F = aLb (Fig. 5.5). Such an equation can be converted to a linear form by converting to logarithms, i.e. log F = a + b log L where F = fecundity, L = length of the fish and a and b are constants. The value of b is normally close to 3 although for some elasmobranchs it is nearer 2. Pitcher and MacDonald (1973) have shown that if mean length is used in such an equation for the estimation of a stock the number of eggs is underestimated because small fish have proportionately fewer eggs than large fish.
The constants are not the same for all species and for one species both a and b may alter with time. Also stocks of the same species can be separated from statistically significant differences between values either of a or b in the fecundity-weight and fecundity-length relationships. As the weight of a fish is normally closely proportional to the cube of the length the fecundity-weight relationship is linear for those species for which b approximates to 3. (Fig. 5.5), and is of the form F = aW + b, where W is the weight of the fish. Another way of expressing such differences is by using a ‘fecundity index’ expressed as either fecundity/weight or fecundity/length³. The latter is the better expression because it avoids the variance in weight within a stock during a season caused by growth of the gonads.
Baxter, I.G., 1963 A comparison of fecundities of early and late maturity stages of herring in the north-western North Sea. Rapp.P.-V.Réun.Cons.Perm.Int.Explor.Mer, 154:170–4
Boyar, H.C. and R.A. Clifford, 1967 An automatic device for counting dry fish eggs. Trans.Am. Fish.Soc., 96(3):361–3
Fischer, W. and F. Balboutin, 1970 On the investigation of ovarial cycle and fecundity of fish with special reference to partial spawners. Ber.Dtsch.Wiss.Komm.Meeresforsch., 21:56–77
Ford, E., 1921 A contribution to our knowledge of the life histories of the dogfishes landed at Plymouth. J.Mar.Biol.U.K., 12:468–505
Holden, M.J., 1974 Problems in the rational exploitation of elasmobranchs and some suggested solutions. In Sea fisheries research, edited by F.R. Harden-Jones, London Elek Science, pp.117–37
Holden, M.J., The fecundity of Raja clavata in British waters. J.Cons.Int.Explor.Mer, (in press)
Howard, G.B. and A. Landa, 1958 A study of the age, growth, sexual maturity, and spawning of the Anchoveta (Cetengraulis mysticetus) in the Gulf of Panama. Bull.Inter-Am. Trop.Tuna Comm., 2:391–467
Macer, C.T., The fecundity and breeding biology of the horse-mackerel (Trachurus trachurus) in British waters. J.Fish.Biol., (in press)
Parrish, B.B. and I.G. Baxter and M.J.D. Mowat, 1960 An automatic fish egg counter. Nature, Lond., 185(4715)777
Pitcher, T.J. and P.D.M. Macdonald, 1973 A numerical integration method for fish population fecundity. J.Fish.Biol., 5:549–53
Schaefer, M.B. and C.J. Orange, 1956 Studies on the sexual development and spawning of yellow-fin tuna (Neothunnus macropterus) and skipjack (Katsuwonus pelamis) in three areas of the eastern Pacific Ocean by examination of gonads. Bull.Inter-Am. Tuna Comm., 1(6):281–349
| Stage | State | Description |
|---|---|---|
| I | Virgin | Sexual organs very small, situated close to vertebral column. Testis and ovary transparent, colourless or grey. Eggs not visible to naked eye. |
| II | Maturing virgin | Testis and ovary translucent, grey- red. Length of gonads 1/2, or slightly more, of length of ventral cavity. Individual eggs can be seen with magnifying glass. |
| III | Developing | Testis and ovary opaque, reddish with blood capillaries. Occupy about 1/2 of ventral cavity. Eggs visible to naked eye as whitish granular material. |
| IV | Developed | Testis reddish-white, no milt produced under pressure. Ovary orange-red. Eggs clearly discernible, opaque. Testis and ovary occupy about 2/3rds of ventral cavity. |
| V | Gravid | Sexual organs fill ventral cavity. Testis white. Drops of milt produced under pressure. Eggs completely round, some already translucent and ripe. |
| VI | Spawning | Roe and milt run under slight pressure. Most eggs translucent with few opaque eggs left in ovary. |
| VII | Spent | Not completely empty, no opaque eggs left in ovary. |
| VIII | Resting | Testis and ovary red and empty. A few eggs in state of resorption. |
| Stage | State | Description |
|---|---|---|
| I | Immature | Ovary and testis about 1/3rd length of body cavity. Ovaries pinkish, translucent; testis whitish. Ova not visible to naked eye. |
| II | Maturing virgin and recovering spent | Ovary and testis about 1/2 length of body cavity. Ovary pinkish, translucent; testis whitish, more or less symmetrical. Ova not visible to naked eye. |
| III | Ripening | Ovary and testis is about 2/3rds length of body cavity. Ovary pinkish-yellow colour with granular appearance, testis whitish to creamy. No trans- parent or translucent ova visible. |
| IV | Ripe | Ovary and testis from 2/3rds to full length of body cavity. Ovary orange-pink in colour with conspicuous superficial blood vessels. Large transparent, ripe ova visible. Testis whitish- creamy, soft. |
| V | Spent | Ovary and testis shrunken to about 1/2 length of body cavity. Walls loose. Ovary may contain remnants of disintegrating opaque and ripe ova, darkened or translucent. Testis bloodshot and flabby. |
Fig. 5.1 Sex differentiation in plaice- above: male; below: female
Fig. 5.2a Automatic fish egg counter
Fig. 5.2b Automatic fish egg counter
| Fig. 5.3 | Maturity stages of male elasmobranchs;
mixopterygia are cross hatched. (a) immature, (b) maturing and (c) mature |
| Fig. 5.4 | Maturity stages of female elasmobranchs (a) immature, (b) maturing, (c) mature oviparous species, (d) mature live-bearing species (wall of oviduct highlyvascularised). |
Fig. 5.5 Fecundity/length and fecundity/weight relationships in herring
