Version classiqueVersion mobile
OpenEdition Books

The inland water fishes of Africa

 | 
Didier Paugy
, 
Christian Levêque
, 
Olga Otero

Growth and ontogeny

Christian Levêque

Résumé

Growth is one of the most complex processes for an organism. On the metabolic level, part of the energy consumed will be devoted to increasing its weight, but the proportion of energy used to generate living matter depends on the age of the individuals, their physiological state, their environmental conditions, etc.

Texte intégral

First stages of development

1Little is known about the first stages of development in African fishes. A review of literature shows that data is only available for 18 of 74 identified families (Cambray & Teugels, 1988).

Ontogeny and main stages of development
Ontogeny is the process of differentiation of the different stages of development in the life of an organism. We usually distinguish several periods in the life of a fish. (Balon, 1981, 1984 and 1986):
• the embryonic period which begins with fertilization and is characterized by exclusively endogenous nutrition from the egg yolk;
• the larval period which begins with the progressive but rapid transition from an endogenous food supply to exogenous feeding. This period is characterized by the presence of temporary larval organisms;
• the juvenile period begins when the fins are well-differentiated and when all temporary organs are replaced by final organs. This stage ends with the first maturation of gametes. This is usually a period of rapid growth sometimes characterized by a specific colouration;
• the adult period begins with the first maturation of gametes. It is characterized by a decrease in somatic growth rate;
• finally, there is sometimes a period of senescence.

2Balon (1985; 1990) distinguishes two broad types of ontogenic trajectories. In the indirect development or “altricial” model, eggs are generally small and produced in large numbers. They yield small, underdeveloped young larvae with only a small volume of yolk that is not sufficient for producing the final phenotype. These young larvae must feed rapidly on small particles to complete their development and are highly vulnerable during this period. The Alestidae Alestes baremoze is a good example of this type of fish (Durand & Loubens, 1971) (figure 11.1), as are Clarias gariepinus (Bruton, 1979a) and Heterobranchus longifilis.

FIGURE 11.1. Early stages of development of Alestes baremoze in the Chad basin (from Durand & Loubens, 1971).

FIGURE 11.1. Early stages of development of Alestes baremoze in the Chad basin (from Durand & Loubens, 1971).

3In the direct development (“precocial”) model, fishes produce a limited number of large eggs with a large amount of yolk, which allows the embryo to develop to an advanced stage. This shortens or eliminates the larval period, and juveniles are already well-formed and thus less vulnerable when they start seeking an external diet. Labeotropheus, a Cichlidae mouth-brooder from Lake Malawi, is a good example of this type of fish that releases a large juvenile (14% of adult size) only 31 days after fertilization (Balon, 1977) (see box “First stages of development in Labeotropheus”). Cyphotilapia frontosa illustrates an even more advanced style, as embryos already start consuming an external diet even while in the buccal cavity where they still have a yolk reserve (Balon, 1985).

First stages of development in Labeotropheus (from Balon, 1977)
Labeotropheus is an endemic Cichlidae in Lake Malawi. It is a mouth-brooder for eggs and juveniles.
Just after spawning, 30 to 50 oval-shaped eggs (3 mm in diameter and 4.4 mm long) are immediately collected in the mouth by females.
They are fertilized in the oral cavity of the female which is almost completely filled by eggs.
The eggs are incubated for 6 days before hatching. At birth, juveniles have a large yolk sac and stay in the oral cavity where they develop and acquire their main morphological structures (skeleton, fins, etc.). The yolk sac is absorbed 21-22 days after fertilization, but embryos stay in the oral cavity of the female up to the 31st day. At this age, they are very well-formed, autonomous, large (approximately 15 mm, i. e., 14% of the adult size) and begin their exogenous foraging.

4There are, of course, intermediate modes between the two broad styles discussed above. Pollimyrus isidori (table 11.I) is an example.

TABLE 11.I. Comparative characteristics in the development of some African fish species.

TABLE 11.I. Comparative characteristics in the development of some African fish species.

References. 1: Durand & Loubens, 1971; 2: Cambray, 1985; 3: Bruton, 1979; 4: Legendre & Teugels, 1991; 5: Kirshbaum, 1987; 6: Balon, 1977; 7: Balon, 1985. Gonado-somatic index (GSI)

Growth estimation

5Growth can be estimated using changes in size or biomass in a given time interval. It corresponds to the organism’s energy intake that is not used for maintaining metabolism.

6There is a great deal of literature on growth estimation methods, and researchers agree that it is difficult to determine the age of tropical fish. Marks on bony structures are harder to read and interpret than on temperate fishes (Mérona et al., 1988). Nonetheless, certain investigations have shown that growth marks generally coincide with drops in temperature, which often occur during low water periods. In some cases, growth marks can be associated with gonad maturation which generally follows a long period of fasting. Formation of marks on bony structures is thus the result of various physiological disturbances, and it is necessary to determine which ones are at play for a given species.

7Determination of growth and age were long considered key elements for models of fish stock dynamics. As a result, a large number of growth curves were calculated for African fishes (Mérona et al., 1988), but most did not attempt to account for the relationships between growth and other biological parameters. Consequently, many of these results are of limited use to understand species biology.

Growth models

8Annual growth can be described as an asymptotic curve, and the most widely used is von Bertalanffy’s model (1938). It is based on bioenergetic principles, with the hypothesis that growth rate is equal to the difference between the anabolic rate and the catabolic rate. The von Bertalanffy model is expressed as the equation:

9Lt = L∞ {1-e-K(t-t0)}

10where L∞ is the asymptotic length, Lt fish length at age t, K is a constant that describes the rate of growth, and t 0 is the hypothetical time at which size is zero.

11Similarly, fish weight at time t is given by the equation:

12Wt = W∞ {1-e-K(t-t0)}

13While von Bertalanffy’s model provides us with a smooth annual growth curve, growth is not a continuous process. At times there are long periods in a year during which growth stops or is sharply reduced. In other words, growth can take place during rather short periods. For instance, 75% of the growth of young Oreochromis andersoni and O. macrochir in the Kafue River occurs during the six weeks of peak water levels (Dudley, 1972). During the dry season, growth is very slow but does not stop (Kapetsky, 1974).

14Welcomme & Hagborg (1977) proposed a growth model for fishes in flood zones, which takes into account a rapid initial increase in length followed by a slower growth period. In the formula Lt+t’= Lt + G (et’), the values of Lt for successive years are the same as the ones provided by the von Bertalanffy model, but growth during the year is calculated for each week. This gives a growth curve that is a more accurate representation of what happens on the field.

15Growth in length and weight gain are not always correlated. Weight gain includes the creation of fat reserves, for instance, or gonad development, phenomena that do not necessarily involve an increase in length (table 11.II).

TABLE 11.II. Some growth data about African fish (from Mérona et al., 1988.

TABLE 11.II. Some growth data about African fish (from Mérona et al., 1988.

S: sex (m: male, f female, t: sexes blended);
L: length (TL: total length; SL: standard length; FL: fork length);
M: size at first maturation; A: average longevity (years);
AMS: average maximal size; MSO: maximal size observed;
L, Kand t0: parameters of the von Bertalanffy equation.

Length-weight relationships and coefficient of condition

16The relationship between fish length and weight takes the form:

17log W = log a + b log L

18where W is the weight, L the length, a and b are constants.

19When growth is isometric, that is, when the fish shape does not change with growth, b is equal to 3. A value below 3 indicates that weight gain is inferior to increase in length. The opposite holds true if b is above 3 (table 11-III).

TABLE 11-III. Length-weight relationships of some African fish species.

TABLE 11-III. Length-weight relationships of some African fish species.

Constant values: a and b.

20The index K, also called the coefficient of condition, is the relationship between body weight (W in g) and length (L in cm):

21This coefficient does not change with length if growth is isometric, which is rarely the case. In many species, however, there may be a variation in body shape or weight during growth, depending for instance on the reproductive cycle or the availability of food. The coefficient of condition makes it possible to monitor the fish’s weight. It is thought to be a good instrument for comparing the overall physiological state of populations during a seasonal cycle or between basins with different ecological conditions.

Longevity, size and growth

22An important characteristic of fishes is that growth never stops during their lifetime but only slows down with age or size, unlike in mammals and birds. One thus cannot associate an absolute average size to a species and, instead, the maximum size observed (MSO) is most often mentioned. Another characteristic is the great variability in growth at an individual level as well as at the level of populations living in physical systems with different ecological characteristics (food, space, temperature, competition, etc.).

23We can posit that the larger a fish, the older it is. If this is the case, there should be a relationship between characteristic growth parameters and a parameter expressing maximum age or size. This indeed holds true, and there is a significant relationship between the maximum size observed (MSO) in wild populations and estimated lengths at 1 year (L1) and 2 years (L2) (Legendre & Albaret, 1991) (figure 11.2). They are, respectively, L1 = 0.205 MSO + 45.57 and L2 = 0.333 MSO + 63.86 for 58 value pairs. These models have been validated experimentally using 6 farmed species.

FIGURE 11.2. Relationship between the maximum size observed (MSO in mm) and the length at 1 and 2 years old for various freshand brackishwater African fish species. (data from Legendre & Albaret, 1991).

FIGURE 11.2. Relationship between the maximum size observed (MSO in mm) and the length at 1 and 2 years old for various freshand brackishwater African fish species. (data from Legendre & Albaret, 1991).

24These results show the existence of a strong positive correlation between the MSO and lengths L1 and L2, which indicates that fishes with a high MSO have a higher growth rate than fishes with a lower MSO.

25The ratios L1/MSO and L2/MSO indicate that African fishes reach about a third and then half their maximum size at the end of the first and the second year, respectively (Legendre & Albaret, 1991). These observations agree with the results obtained by de Mérona (1983) who underscored a similar relationship between the asymptotic length, L∞, and the coefficient K of the von Bertalanffy equation: log K = 2.186–1.048 log L∞ which can be simplified to K = 153/L∞. In order to apply this relationship to field data, de Mérona (1983) also established an empirical relationship between L∞ and the maximum average size (MAS) of the population, which is the average value (or mode) of the largest individuals observed in the population: L∞ = 1.248 MAS. The combination of these two ratios allows a rough estimate of the growth curve of African fishes (Mérona et al., 1988), which serves as a quick model for estimating growth. Data available for African fishes appear to show that fishes in tropical regions do not grow faster than those in temperate zones, contrary to common belief, but this would merit further investigation.

26If the above results highlight general trends, it should be nonetheless recalled that there is high inter-species variability as well as large differences between populations of the same species living in different hydrographic basins. That said, the MSO remains a practical tool for rapid assessment of growth rates in the absence of other data.

Size and age at first maturity

27Puberty is an important event in the life of an animal, and is accompanied by a reduction in somatic growth. Sexual maturation is costly in terms of energy. The reserves accumulated up to that point are used for gonad development, gamete production, development of secondary sex characteristics, etc.

28As a result, there are potential conflicts in energy use between:

  • continued growth of the individual, who could then produce more eggs;
  • immediate production of descendants, which is another type of demographic strategy.

29The age at which the fish reaches sexual maturity thus has evolutionary implications. In stable populations, it is believed that the age at first reproduction evolved in order to maximize reproductive success throughout the fish’s lifetime. The average size at first reproduction, defined as the size class in which at least 50% of the individuals are mature, is often mentioned. This average size may vary depending on the type of physical system and may not be the same for both sexes (see table 11. II).

30Do fishes reach sexual maturity at a given age or size? This question was the object of much discussion but is probably not highly relevant after all, as fishes mature within an age or size range that is probably defined by the combined influence of genetics and environment. For some cichlid species living in different physical systems, growth and age or size at first maturation, as well as MSO, can also be very different (Noakes & Balon, 1982). For O. niloticus, the size at first maturation may vary from 140 mm to 390 mm in the natural environment, and for C. zillii and O. mossambicus, this size can go from 70 and 45 mm respectively in an aquarium, and 135 and 300 mm in a lake or reservoir (table 11.IV).

TABLE 11.IV. Comparison of size at sexual maturity, maximum size observed and longevity for different populations of cichlids. Data from Noakes & Balon (1982) for C. zillii and O. mossambicus; from Trewavas, (1983) for O. niloticus.

Species

Locality

Size at maturation (mm)

Maximal size (mm)

Longevity (years)

Coptodon zillii

Lake Kinnereth

135

270

7

Egypt, ponds

130

250

Lake Naivasha

90

aquarium

70

Oreochromis mossambicus

Lake Kariba

300

390

8

South Africa

390

11

Sri Lanka

150

340

Egypt

300

7

Hong Kong

165

310

4

aquarium

45

Oreochromis niloticus

Lake Turkana

390

640

Lake George

280

400

Lake Albert

280

420

Lake Edward

250

360

Lake Baringo

180

360

Egypt

200

330

9

Buhuku lagoon

140

260

31Size and age at maturation are thus influenced by the environmental conditions in which the fish grows. Thus, most Tilapias are known for their capacity to reproduce at a smaller size when they are in an environment that does not fully suit them. For four wild populations of O. mossambicus living in small pools of water with different sets of environmental characteristics, the age at first maturity varies from 1 to 3 years, and the size at first maturity ranges from 118 to 236 mm for females (James & Bruton, 1992) (table 11.V).

TABLE 11. V. Biological traits of four South African populations of O. mossambicus in different types of habitats (from James & Bruton, 1992).

TABLE 11. V. Biological traits of four South African populations of O. mossambicus in different types of habitats (from James & Bruton, 1992).

32For Cichlidae, changes in the size at first maturity (SFM) are also influenced by fishing pressure. In Lake George, for instance, intensive fishing is thought to have led to the decrease in the SFM of Oreochromis niloticus from 27.5 cm TL in 1960 to only 20 cm in 1972 (Gwahaba, 1973). Likewise, in Lake Turkana, the SFM of O. niloticus went from 39 cm in the 1950s to only 29.6 cm in the 1970s (Harbott & Ogari, 1982).

33There are many other examples of variations in age or size at first maturity for a single species. But the reasons for the shift from a demographic tactic favouring growth to another one favouring early reproduction have not yet been identified in a convincing manner. Diet-related factors are often mentioned. In fact, field data have shown that Tilapias in poor physiological condition have a strong tendency to reproduce at a smaller size than those in better condition. Such a phenomenon has been observed for Oreochromis niloticus in different African lakes (Lowe, 1958), and for Sarotherodon melanotheron in the lagoons of Côte d’Ivoire (Legendre & Écoutin, 1996) (figure 11.3).

FIGURE 11.3. Relationships between the mean weight of a 200 mm

FIGURE 11.3. Relationships between the mean weight of a 200 mm

TL specimen and the size at first maturity (from Legendre & Écoutin, 1996). TL: total length; FL: fork length. Oreochromis niloticus populations from several African aquatic environments (data from Lowe, 1958). Sarotherodon melanotheron from Côte d’Ivoire (from Legendre & Écoutin, 1996).

Dwarf populations

34The existence of dwarf populations that reproduce at much smaller sizes than those customarily observed has also been found in non-cichlid fishes.

35In Lake Chad, Brycinus nurse dageti is a dwarf population of Brycinus nurse nurse and the two species coexist in the lacustrine environment (Bénech & Quensière, 1985). Likewise, in Lake Turkana, Brycinus nurse nana is also a dwarf population of B. nurse (Paugy, 1986). There are also dwarf populations of Schilbe niloticus in Lake Chad (Bénech & Quensière, 1985) and of Ethmalosa fimbriata in some bays of the Ébrié lagoon in Côte d’Ivoire (Albaret & Charles-Dominique, 1982). These phenomena have not been well explained to date. The cline detected in Schilbe mandibularis (Lévêque & Herbinet, 1982) that leads to dwarf populations in Liberia has not been explained either.

36A distinctive aspect of dwarfism is that of “stunted” populations: fishes are not small for their age, but old for their size (Noakes & Balon, 1982). These populations have an accelerated ontogeny which manifests as very early sexual maturity. Thus, Coptodon zillii which usually reaches sexual maturity at around 2 or 3 years and a size of 200-300 mm can become mature in a few months at a size below 100 mm when it is introduced into small pools (Fryer & Iles, 1972). In a hot spring, Hecht & Zway (1985) observed a population of O. mossambicus in which the largest individual was a male measuring 105 mm SL, while the smallest mature female measured 35 mm SL at one year of age. This “stunting” phenomenon has been observed in nature when shallow lakes dry out, or when small pools are isolated from the main water mass. In Lake Turkana, where O. niloticus is mature at an average size of 280 mm TL, it has been observed that populations living in springs isolated from the lake reproduced at 80 mm TL (Harbott & Ogari, 1982). Here, the mechanisms at play are not well-identified either.

Environmental conditions and growth characteristics

37Most species have developed a type of demographic strategy matching a habitat that is favourable to them, as they live there and have adapted to it. Nonetheless, when some species find themselves in worse conditions, they can modify their reproductive behaviour, sometimes within large margins.

38The phenotypic plasticity of Oreochromis mossambicus, which is found in a large variety of habitats, is well known. Its level of investment in the protection of eggs and young varies depending on the conditions of the physical system. By studying four populations living in physical systems that differed greatly in their abiotic and biotic characteristics, James & Bruton (1992) showed that the species is able to shift from its usual mode of direct development to an indirect one when the conditions of the habitat are more drastic. The populations of the Kowie lagoon and of Mill Farm lake become mature at a length around 70% of their maximum size, which is typical of Cichlidae living in stable physical systems and who have an indirect mode of development. The population in the Rufanes pool, meanwhile, showed early maturity and smaller eggs, characteristics typical of an indirect mode of development. Populations in Rufanes and Lake Bradshaw, habitats presenting more difficult conditions, are much more fecund than the others. Generally speaking, when all the biological characteristics are taken into account, there is a shift from a direct development style in Mill Farm lake to an indirect one in Rufanes pool.

Table des illustrations

Titre FIGURE 11.1. Early stages of development of Alestes baremoze in the Chad basin (from Durand & Loubens, 1971).
URL http://books.openedition.org/irdeditions/docannexe/image/25214/img-1.jpg
Fichier image/jpeg, 243k
Titre TABLE 11.I. Comparative characteristics in the development of some African fish species.
Légende References. 1: Durand & Loubens, 1971; 2: Cambray, 1985; 3: Bruton, 1979; 4: Legendre & Teugels, 1991; 5: Kirshbaum, 1987; 6: Balon, 1977; 7: Balon, 1985. Gonado-somatic index (GSI)
URL http://books.openedition.org/irdeditions/docannexe/image/25214/img-2.jpg
Fichier image/jpeg, 179k
Titre TABLE 11.II. Some growth data about African fish (from Mérona et al., 1988.
Légende S: sex (m: male, f female, t: sexes blended);L: length (TL: total length; SL: standard length; FL: fork length);M: size at first maturation; A: average longevity (years);AMS: average maximal size; MSO: maximal size observed;L, Kand t0: parameters of the von Bertalanffy equation.
URL http://books.openedition.org/irdeditions/docannexe/image/25214/img-3.jpg
Fichier image/jpeg, 181k
Titre TABLE 11-III. Length-weight relationships of some African fish species.
URL http://books.openedition.org/irdeditions/docannexe/image/25214/img-4.jpg
Fichier image/jpeg, 369k
Légende Constant values: a and b.
URL http://books.openedition.org/irdeditions/docannexe/image/25214/img-5.jpg
Fichier image/jpeg, 383k
URL http://books.openedition.org/irdeditions/docannexe/image/25214/img-6.jpg
Fichier image/jpeg, 11k
Titre FIGURE 11.2. Relationship between the maximum size observed (MSO in mm) and the length at 1 and 2 years old for various freshand brackishwater African fish species. (data from Legendre & Albaret, 1991).
URL http://books.openedition.org/irdeditions/docannexe/image/25214/img-7.jpg
Fichier image/jpeg, 84k
Titre TABLE 11. V. Biological traits of four South African populations of O. mossambicus in different types of habitats (from James & Bruton, 1992).
URL http://books.openedition.org/irdeditions/docannexe/image/25214/img-8.jpg
Fichier image/jpeg, 107k
Titre FIGURE 11.3. Relationships between the mean weight of a 200 mm
Légende TL specimen and the size at first maturity (from Legendre & Écoutin, 1996). TL: total length; FL: fork length. Oreochromis niloticus populations from several African aquatic environments (data from Lowe, 1958). Sarotherodon melanotheron from Côte d’Ivoire (from Legendre & Écoutin, 1996).
URL http://books.openedition.org/irdeditions/docannexe/image/25214/img-9.jpg
Fichier image/jpeg, 174k

Auteur

Research director at IRD. He has published numerous works on the systematics, distribution, and ecology of West African fresh water fishes, and expanded the collections of the MNHN in Paris.

© IRD Éditions, 2017

Conditions d’utilisation : http://www.openedition.org/6540

Lire

Accès exclusif

open access

Offert par L’éditeur de ce site