Suplement XXVIII

Pakistan Journal of Science, 1950, 3(1) : 39-40.

 Spawning behaviour and early stages
 in the Development of a larvivorous
 fish Colisa lalia Hamilton.

Utility of Colisa lalia as a larvivorous fish has been demonstrated in the field and in the laboratory by experiments and observations extending over a period of twelve years (Hamid Khan 1943). This little fish, not more than 3 inches in length, breeds successfully in ponds and in aquaria, and is a „bubble-nest builder.”

The first account of nidification of Colisa (Trichogaster) fasciata was given by Paul Carbonnier (1875). Innes (1942) gives a detailed account of nest building in the family Anabantidae to which Colisa belongs. „The outstanding feature of the breeding of this family of fishes is the floating nest of bubbles which they construct and in which the eggs are placed, hatched and the young tended. These bubbles are formed by the male as he comes to the surface, draws a little air in his mouth and envelops it in a filmlike saliva. When released, the globule naturally floats to the surface. Endless repetition of this act piles up what looks like a little mound of very fine soap-bubbles,” (Innes. 1942). As there are no salivary glands in fish, by 'saliva’ Innes probably means mucus secreted by mucous cells present in the buccal cavity.

The present study is the result of observations recorded in spawning behaviour of Colisa lalia, kept in aquaria in the Laboratory and in ponds in the Punjab. During winter male is not easily distinguishable from female. But in April and May it starts assuming its nuptial colour of red and blue stripes on its body and fins. The fish, however, spawn in July and August, temperature of water ranging from 74°F. to 80°F. One pair spawned as late as October.

It has been observed that in an aquarium with large number of male and female of Colisa lalia, nest is built by a male during the day, but is destroyed during the night by the other jealous males. It is best to separate a pair at the approach of breeding season and keep it in an aquarium. The male immediately afterwards starts constructing the nest, invites the female to the nest, where the eggs are laid by the female and placed in the nest. An interesting case was observed when a ripe female was introduced in an aquarium having a spent male, the female chased the male which ran away whenever the female approached. The spent male was taken  out and a mature male was introduced in that aquarium. The reaction was reverse, the male chased the female. But after a time they settled down, the male constructed the nest next day and invited the female to lay the eggs.  Colisa lalia incorporates bits of plants in the nest such as fine leaves of Hydrilla (Hamid Khan, 1943). The nest measures 3.4 inches to 4 inches in diameter and 2 inches above water surface. In one case the nest was not so compact and measured  6 inches in diameter. Each nest contains 100 to 400 eggs.

After spawning the male takes charge of the nest and drives the female away. If the female is not removed from the aquarium, it is likely to be killed by the male. The male keeps the bubble-nest compact by producing more bubbles after the eggs are laid. The egg measures 1 mm. in diameter and has a large amount of yolk with an oil globule which keeps the egg buoyant.

Outlines of the embryo become defined 7 hours after fertilization. Two hours later eyes are visible, head end is swollen and there are 21 somites on the body. Twelve-hour stage shows 27 somites, heart has made its appearance, pigment cells are also present and eyes are prominent, wrigling movement has started. Otocyst is visible. The blood circulation is simple. The eggs hatch out in about 26 hours. Just before hatching eyes have become prominent but without pigment, and yolk circulation is established. Four hours after hatching pectoral fin has appeared as a lobe. Tail is elongated and blood circulation has extended almost to the tail. In some cases three fin rods are visible in the Caudal fin. Fifteen hours after hatching, yolk sac is much reduced, pectoral fins are well developed and heart is Three Chambered. Forty-eight hours after hatching, yolk sac is almost completely absorbed, mouth opens for respiration, gills are functioning, simple gut is formed, but anus is not yet open.

Sixty hours after hatching, yolk is totally absorbed, anus is open, and food is in the simple tubular gut.  Pectoral fins are well developed. The nest has disappeared and the little ones  start moving about. At this time the male, as also been reported by the previous authors (Innes, 1942) starts eating the little ones. It becomes necessary to remove the male at this stage to save the progeny. In wild state the little ones protect themselves by concealing themselves in the weeds.

Curiously enough Colisa has almost completely disappeared from the pond where they used to breed in large number. At one time 293 nests were counted in one day in July. This disappearance seems to be correlated with the absence of Hydrilla, which has also disappeared from that pond. In a pond where Colisa is introduced, Hydrilla appears to be a necessity not only for the nest building but also for the protection of young ones.


Zoology Laboratory
University of the
Panjab, Lahore,


Innes, W. T. (1942) Exotic Aquarium Fishes. Innes Publishing Co.. Philadelphia, U.S.A.

Khan, H. (1943). „On the relative value of certain Larvivorous fishes from the from the Punjab, with notes on their habits and Habitats.” Ind. Journ. Vet. Sci. 13-4-1943.


Pakistan Journal of Science,  pages 78-81

On the distribution of taste buds in certain teleost fishes

„””A knowledge of the senses and instincts of fishes is necessary for making Fisheries a success. Feeding habits and the sense of taste are, in this respect, perhaps, as important as any other aspect of fish-life, because without a knowledge of these it is not possible to manipulate economically such processes as the 'stocking of tanks’.

Anatomical studies on the organs of taste in fishes date back to 1827 when Weber observed the taste-buds on the peculiar palatal organs of Carp. Taste- buds, or the terminal-buds, as they were called in this case, were observed in the outer skin of fishes by Leydig (1851). A strict distinction between the terminal- buds and the 'neuromasts’ or the organs of the lateral line system, was first established by Schulze (1870). Merkel’s (1880) interpretation of both the terminal buds and the neuromasts as the orgns of touch created a great deal of confusion which was finally resolved by the studies  of Herrick (1902) who confirmed the conclusions of Schulze and showed that the very nerve-supply of the two systems was different.

A typical taste-bud consists of a bundle of sensory cells resting on a papilla of corium or submucosa  and accompanied by the supporting or sustentacular cells. A small pore in the epidermis, called the sensory pore, keeps the refractive processes of the sensory cells in  communication with the outside. Typical taste-buds have been described by various authors working on the histology of the skin (Bhatti, 1938) or the alimentary canal (Dawes, 1929, Rogick, 1931, Al- Hussaini, 1949). Except for a brief reference by Al-Hussaini (1949) no attempt has been made to correlate the number or localisation of taste-buds with the nature of food or with feeding habits.

The present study was undertaken to find out the structure and distribution of taste-buds in some fresh-water teleost fishes and to correlate the findings with the feeding habits. The fishes selected for this study are Rita rita Ham., an insectivorous fish which feeds also on molluscs;  Ophicephalus gachua a piscivorous fish; Cirrhina mrigala Ham., a bottom-feeder herbivorous fish and Colisa (Trichogaster) lalia Ham., a larvivorous fish.”””

„””Colisa lalia. — Taste-buds are present in Colisa lalia in the skin, buccal cavity and the pharynx, with their relative frequency increasing in the same order, becoming quite numerous in the pharyngeal lining. In the buccal cavity the taste-buds are almost as tall as broad. The number of sensory cells in these taste-buds is large and their refractive processes remain more or less parallel. In pharynx taste-buds of two shapes are met with. Some taste-buds are globular with their refractive processes apparently meeting in a point. The number of sensory cells in these taste-buds is quite large. Other taste-buds are flask-shaped, with a round body and a long neck formed by refractive processes.Taste- buds of shapes somewhat in between the two types described are also met with.”””

„””Colisa lalia is a larvivorous fish and seems to have a taste for mosquito larvae and Crustacea. Like Rita rita, this fish has its sight subordinated to its sense of taste so far as the seeking of food is concerned. Experiments on the feeding of this fish were made by Hamid Khan (1943). He found that these fish, when given charcoal pieces of the size of mosquito larvae, would readily eat them, but would immediately spit them out. These observations are substantiated by the distribution of taste-buds in this fish. „””

Hamid Khan

Zoology Laboratory, Ahsanul Islam
University of the Panjab, Lahore.


J. Annamalai Uni., Sei., 27, 81-89 (1966-69)

— 81—




Department of Zoology, Annamalai University, Annamalainagar


A detailed account of the biology of Colisa lalia distributed in the freshwaters of different regions is found in the works of Innes, and Mellan and Lainer (quoted by Jones, 1946), Axelrod and Schultz (1955), Axelrod (1956), Forselius (1957), Mookerjee and Mazumder (1960) and Abraham (1962). The present study subscribes further to the knowledge on the biology of this species confined to the waters of Annamalainagar. Also, it is considered that a knowledge on this aspect would be desirable for a study undertaken by one of us (V.G.K.) on the morphology, development and secretory activity of the corpuscles of  Stannius in this species (Krishnamurthy, 1964, 1967, 1968).

Material and Methods

The fry of C. lalia were collected from the boating canal, Annamalainagar during July – August, 1961 and February – March, 1962 and reared separately in cement tanks for observations of their growth. These tanks were provided with plants such as Vallisneria sp., Hydrilla sp., Lemna minor and Myriophyllum sp.

The fry at the time of catch measure each 6 mm in length. About 100 fry were transferred every fortnight from the stock to a petri-dish filled with water; and the total length of each fry was measured with a paraffin coated centimeter checker paper fixed to the bottom of the dish.

After recording the measurements the fry were returned to the tanks. The adults were  taken out of water and measured  with a metric scale; while measuring they were able to withstand the exposure to air, since they breathe by

— 82—

suprabranchial organs. The mean values of growth during fortnightly intervals were calculated separately for the summer-bom and monsoon-born fishes and the respective curves were drawn for these values. A few of the fry from each sample after removal of their head and tail were fixed in Bouin’s fluid. These were decalcified in formol-nitric (10: 1.5 v/v) and after usual procedure embedded in paraffin wax. Deparaffinised sections, 8 microns in thickness, stained in Delafield’s haematoxylin-eosin and Heidenhain’s haematoxylin were used to examine the condition of the gonad in relation to the length of the fish.


Day (1958) considers that Colisa (Cuvier and Valenciennes), the generic name, is a  synonym for Trichogaster (Bloch and Schneider). Smith (quoted by Abraham, 1962) is of the opinion that all the Indian species belong to the genus Colisa and all the Southeast Asian species belong to the genus Trichogaster. Colisa one of the seventeen genera of the family Anabantidae, suborder Labyrinthici, includes four species namely, chuna, fasciata, labiosa and lalia. C. lalia originally inhabited the Ganges and Jumna rivers (Forselius, 1957). It spread far south and till 1944. it was confined to north of the Krishna district especially in the Godavari waters, when the Madras Fisheries Department stocked this species in the Madras waters. Since then it has established itself widely in almost all the freshwaters of the Madras State including the uphill waters of Ooty and Yercaud (Abraham. 1962).

This species is reported to grow to a maximum length of 60 mm. It exhibits chromatic sex dimorphism. The males have a mosaic of red and green alternating bands on a blue-grey background. The colour especially is males is intensified in a typical anabantid manner during breeding seasons. It is a hardy and peaceful fish excepting while breeding. It tolerates a temperature ranging between 65° and 85°F. Higher temperatures of the water induce brilliant colouration and accelerate spawning of the fish. This species is considered to be a perennial breeder (Abraham,

— 83—

1962). The existing reports on its breeding habits are based on the observations made in laboratories. Little is known about its spawning periodicities and breeding habits under natural conditions (Mookerjee and Mazumder, 1960). It builds a bubble nest over bits of plants dead and alive in which operation the female assists her mate. There is a prolonged courtship before and after nest building; and the pair rests for a while after ejaculation and oviposition. The male recovers first and collects the deposited eggs in its mouth and blows them into the nest. Simultaneously the male drives off the female from its vicinity and guards the nest.


The fry of C. lalia are available only twice a year i. e., during February-March and July-August. Late  April or September practically no fry  is caught and this indicates that this species possesses two distinct peaks of sexual activity. During late summer i. е., in May adult fishes possess fully mature ovary but they do not spawn; their oocytes get entirely resorbed.


Fry measuring about 38 mm and below in length do not show any secondary sexual characters. It is only beyond this length the male fry are identified by their secondary sexual colouration. Therefore the growth is estimated up to 38 mm for both sexes combined; and over 38 mm it is estimated for males and females separately (Figs. 1, A-D). It is found that the males grow faster than the females exceeding in their length by 1.6 mm during summer (Fig. 1, A) and 2.6 mm during monson (Fig. 1, B; Table I).

Moreover, the monsoon-born fishes grow rapidly during their adult stage and reach their maximum length in summer while the summer-born fishes though they grow faster during their fry stage, show a lower rate of growth during monsoon and reach their maximum length about four weeks after the normal period required of the monsoon-born fishes for reaching their maximum growth (Figs. 1 C and D).

— 84—

Table I: Mean values of body length (mm) of different
growth stages of Colisa lalia during monsoon and
summer (1961-1962)

Body length of individuals
Period of growth      Before attainment of chromatic sex dimorphism
in bi-weeks              Monsoon (1961)              Summer (1962)

.                                                                              At time of catch

1                    11.8                                13.2
2                    18.0                                19.8
3                    23.5                    Pre-adult 25.0
4       Pre-adult 28.5                                30.4
5                    32,0                                34.6
6                    34.0                         Adult 36.4
7              Adult 35.6                               37.2
8                     36.8                                38.0


After attainment of chromatic sex dimorphism

Growth during                      Growth during
summer (1962)                      monsoon (1962)

Male        Female                   Male      Female

 9        39.2         38,6                    38.6        38.0
10       41.0         39.2                    40.0        38.8
11       41.8         40.0                    44.8        39.4
12       42.6         40.6                    44.6        41.6
13       43.6         41.4                    46.2        43.4
14       44.0         42.2                    47.6        45.4
15       44.8         42.8                    48.2        46.2
16       45.8         44.0                    48.4        46.6
17       46.8         44.8                    48.6        47.0
18      47.4         45.2                        —           —
19       48.2         45.6                        —           —

— 85—

Fig. 1:  Growth in body length of Colisa lalia during the monsoon and summer seasons of 1961 – 1962

. .. . .. .. . growth before attainment of secondary sex coloration;
— • — • — growth after attainment of secondary sex coloration
mg – monsoon growth; sg – summer growth.

— 86—


The primordial germ cells appear as a layer between the mesoderm and endoderm and lie ventrolateral to the kidney in fry measuring 6 mm in length (Fig. 2). They multiply, descend as a mass suspended by the mesentery and occupy a positionbesides the intestine. This mass of primordial germ cells along with stroma cells forms the rudiment of the gonad and becomes easily distinguished in fry 12 mm long. The primordial germ cells divide to form primary oogonial cells; and these in turn divide forming the secondary oogonial cells (Figs. 3 and 4). The secondary oogonial cells transform into immature oocytes and this transformation is almost completed at 24 mm stage of the fry. The ovary now is packed with uniform immature oocytes (Fig. 5). The immature oocytes grow and become mature in about 35 mm long fishes.

 Table 2 : Stages of development of the ovary and
the corresponding body growth of female* C. lalia

Stages of growth     Body length     Stage of development of
of the fish              mm               the ovary

Fry or juveniles  8 – 24       Organisation of ovary;
formation of immature oocytes

Pre-adults     24 – 34   Stages of growth of immature
oocytes into mature ones.

Adults          35 and above Ovary in mature and spent
condition; contains also oocytes of
next crop.

*Sex identified from the histology of the gonad, though the external features do not indicate any sex difference in fishes upto 38 mm. in body-length.

— 87—



In most of the tropical freshwater fishes spawning takes place principally at the beginning and end of monsoon seasons as a result of the profound physical, chemical and biotic changes taking place in connection with monsoon rains and inundations (Forselius, 1957). Hermes (quoted by Forselius, 1957) during his field studies in central Thailand has observed nests of anabantids mainly Trichogaster species and Trichopsis vittatus specially in the beginning and end of rainy season.

Only a little is known about the breeding habits and spawning of C. lalia under natural conditions. Based on ova diameter, Abraham (1962) has reported that this species is a perennial breeder. In the present study spawning periodicity is determined from the availability of fry. It was referred earlier on that the fry of C. lalia are available only twice a year i. e., during February- March and July-August, and by the close of each season no fry is caught. Further it was observed that after fortyfive days from the time of catch the monsoon-fry grow into pre-adults; while the summer-fry attain the pre-adult stage after a month. From these it is evident that this fish spawns only twice a year, with the result its fry occur only during the periods mentioned above. The absence of these fry during other parts of the year as observed in the present study seems to contradict the observations of Abraham (1962) who reports that this species is a perennial spawner. However, her observation that majority of these fishes spawn around December and again during June-July partly conforms with the two spawning seasons observed in the present investigation.

Mellen and Lainer (quoted by Jones, 1946) have stated that C. lalia spawns several times during summer at 75°F. It is quite robable that higher temperature of the water has accelerated the

— 88—


— 89—

oocytes transform into mature oocytes. The 'adult’ refers to the stage when the first crop of oocytes attains maturity.


Spawning periodicity of Colisa lalia has been determined from the availability of fry under natural conditions.

The growth of fry and adults during summer and monsoon seasons has been critically examined. The growth of this species is interpreted in terms of the growth of the gonad.


One of us (V.G.K.) is grateful to the Ministry of Education, Government of India for the financial assistance.


Abraham, J. G., 1962. Studies on Colisa lalius (Cuv. and Val.). Fisheries Station Reports and Year Book, April 1957-March 1958. Department of Fisheries, Government of Madras.

 Axelrod, H. R., 1956. Tropical fish as a hobby, A guide to selection, care and breeding. George Allen and Unwin Ltd, London.

Axelrod, H. R., and Schultz L. P. 1955. Handbook of tropical aquarium fishes. McGraw-Hill Book Co., New York.

Brody, S., 1945. Bioenergetics and growth. Reinhold, New York.

Das, K. N., and Das Gupta, B. N., 1945. Factors influencing the spawning of Indian carps. Froc. Nat. Inst. Sei. India, 11, 324-327.

Day, F., 1958. The Fishes India. William Dawson and Sons Ltd., London. s. a.

  — 90—

Forselius, Sien., 1957. Studies of Anabantid fishes I. II and III. Zool. Bidrag Uppsala, 32, 97-597.

Hussain, A., 1945. Factors influencing the spawning of Indian carps. Proc. Nat. Inst. Sci.
India, 11, 320-324.

Jones, S , 1946. Breeding and development of Indian freshwater and brackish water fishes II. J. Bombay Nat- Hist. Soc., 46, 453-472.

Krishnamurthy, V. G., 1964. Corpuscles of Stannius in Colisa lalia (Hamilton – Buchanan). Naturwiss., 51, 344-345.

Krishnamurthy, V. G., 1967. Development of the Corpuscles of Stannius in the Anabantid teleost, Colisa lalia. J. Morph., 123(2), 109-120.

Krishnamurthy, V. G., 1968. Histochemical and biochemical studies of the corpuscles of Stannius of the teleost fish, Colisa lalia. Gen. Comp. Endocrinol., 11, 92-103.

Merriman, D., and Schedl, H. P., 1941. The effects of light and temperature on  gametogenesis in four-spined stickleback, Apeltes quadrants (Mitchill). J. exp. Zool., 88, 413-449.

Mookerjee, H. K., and Mazumder., 1960. On the life history of Colisa lalius (Hamilton). Proc. Zool. Soc., 13, 29-38.

Pickford, G. E., and Atz, J. W., 1957. The physiology of the pituitary gland of fishes. New York Zoological Society, New York.



The review of applied entomology. Series B. Medical and veterinary, Vol, 33, 1945 ,  p. 144

 Hamid Khan. On the relative Value of certain larvivorous Fishes from the Punjab, with Notes on their Habits and Habitats.— Indian J. vet. Sci. 13 pt. 4 pp. 315-325, 23 refs. Delhi, 1944.

The following is mainly based on the author’s summary.

Colisa lalia, Ambassis baculis and Barbus sophore, three larvivorous fish of the Punjab, were observed in captivity to devour 148, 136 and 90 mosquito larvae per fish per day, respectively. A study of their gut contents indicated that  C. lalia and A. baculis invariably feed on mosquito larvae when put into a reservoir where these are present. In the absence of mosquito larvae, they feed on Crustacea, rotifers and the larvae of aquatic insects. Though B. sophore fed actively on mosquito larvae in the laboratory, in ponds it mostly feeds on mud, decayed vegetation, algae and aquatic weeds, and thus reduces the food-supply of mosquito larvae. All three species bred successfully in  reservoir water. Only C. lalia survived in a mixture of equal parts of reservoir water and sewage water, but all could thrive in such stagnant waters as lie near towns and villages. Notes are given on their breeding habits. All need sub-surface vegetation, especially in July and August, to enable them to lay eggs. The investigations made have demonstrated the effectiveness of C. lalia and A. baculis in destroying mosquito larvae in reservoirs and ponds, and their use is recommended. It is calculated that to free water of mosquito larvae as quickly as possible, a population of 2,500 fish per acre of water is necessary.