III/E-5
ON A NEW ARTIFICIAL FEED FOR CARP FRY

by

M.A.V. LAKSHMANAN, D.S. MURTY, K.K. PILLAI and S.C. BANERJEE
Central Inland Fisheries Research Sub-station
Cuttack, India

Abstract

A series of laboratory experiments on rearing of carp fry from hatchling to 15-day old fry were conducted using 19 cheap and easily available items of feed. Besides individual items, the promising items of feed were prepared into combinations of two, three, four and five and in different ratios. Survival and growth of fry were taken as the criteria for determining the relative value of the various items of food.

The data were statistically analyzed in detail comparing the results of all experiments and the conclusion was drawn that a mixture of notonectids (insects), small prawn and cow-pea in the ratio of 5:3:2 yielded most satisfactory results. Any other ratio of the three items did not show comparable result.

The advantages of using the aquatic insect and prawn populations and comparative economy of the supplementary feed are discussed.

Field experiments conclusively proved the feed to be effective for all species of cultivated carps, in giving better production of healthy and well grown fry. The feed appears to satisfy all the essential requirements for commercial use. The mixture of oil cake and rice bran which is the commonly used feed in India, has been found to be less effective.

UN NOUVEL ALIMENT ARTIFICIEL POUR LES ALEVINS DE CARPE

Résumé

Dans une série d'expériences en laboratoire, des alevins de carpe ont été élevés depuis l'éclosion jusqu'à l'âge de 15 jours avec une nourriture artificielle comportant 19 types d'aliments faciles à obtenir et peu coûteux. Ces aliments ont été utilisés isolément ou en rations composées comprenant jusqu'à 5 aliments différents. Les taux de survie et de croissance des alevins ont servi de critères pour apprécier la valeur relative des divers types d'aliments.

En procédant à une analyse statistique détaillée et comparative des données fournies par toutes les expériences, on en est arrivé à la conclusion que les meilleurs résultats étaient obtenus avec un mélange de notonectes (insectes), de petites crevettes et de dolics dans la proportion 5:3:2. Aucune autre formule de mélange de ces trois aliments n'a donné des résultats comparables.

Les auteurs examinent les avantages qu'il y a à utiliser les insectes aquatiques et les crevettes ainsi que l'économie relative de l'alimentation d'appoint.

Des expériences en étang ont nettement démontré qu'avec toutes les espèces de carpes d'élevage, cette formule alimentaire permettait d'obtenir une production améliorée d'alevins sains et bien développés. Cet aliment paraît répondre à toutes les exigences d'une utilisation commerciale. Le mélange de tourteaux d'oléagineux et de son de riz qui est l'aliment artificiel le plus répandu en Inde, donne de moins bons résultats.

NUEVA ALIMENTACION ARTIFICIAL PARA LOS JARAMUGOS DE CARPAS

Extracto

Empleando 19 clases de alimentos baratos y fáciles de obtener, se realizaron una serie de experimentos de laboratorio sobre la cría de jaramugos de carpa desde la fase de alevín recién eclosionado hasta la de jaramugos de 15 días. Además de los distintos tipos de alimentos aislados, se prepararon otros que podrían ser eficaces preparados combinando dos, tres, cuatro y cinco de ellos se adoptaron como criterio para determinar el valor relativo de los distintos productos alimenticios.

Se analizaron estadísticamente los datos con todo detalle, comparando los resultados de todos los experimentos y se llegó a la conclusión de que una mezcla de notonéctidos (insectos), camarones pequeños y caupí en la proporción 5:3:2 daba los resultados más satisfactorios. Cualquier otra proporción de los tres alimentos no dió resultados parecidos.

Se examinan las ventajas de emplear el insecto acuático y las poblaciones de camarones y la economía de la alimentación suplementaria.

Los experimentos de campo demostraron de un modo concluyente que la alimentación servía para todas las clases de carpas cultivadas, dando una mejor producción de jaramugos sanos y bien desarrollados. La alimentación parece que satisface todas las necesidades esenciales para poder usarla en escala comercial. Se ha visto que la menos eficaz es la mezcla de tortas oleaginosas y salvado de arroz, que es el alimento generalmente empleado en la India.

1 INTRODUCTION

Though fish cultural practices are in vogue all over India, the matter of artificial feeding of fish to get enhanced production does not appear to have received the attention it deserves. A review of available literature on this aspect of fish culture seems to indicate that almost all the work done has been concentrated on feeding the fry at the hatchery or nursery stage. Thereafter, availability of food for the growing fry, fingerlings and yearlings is left to the natural food production of the waters.

The obvious reasons for this omission of artificial feeding appear to be the consideration of economy in undertaking this operation in larger waters, as well as lack of sufficient information on suitable supplementary feeds for the various cultured fishes. However, fish culturists such as those in Japan (Kawamoto, 1957) who supplied supplementary diet in one form or another for rearing their stock of fish and obtained high production have sufficiently justified the usefulness of artificial feed in enhancing fish production.

The common items of feed used in India in nursery and rearing ponds are various types of oil cakes mixed with rice bran. With the recent initiation of intensive fish culture programs in the various States of India, it was felt necessary to develop a more suitable and balanced diet for cultured fish. With this in view, preliminary work commenced in this Research Institute, the main aim of which was to enhance fish production within a minimum rearing period. While the present work is limited to the observations on survival and growth of fish at the fry stage, it is proposed to continue the investigations to evolve suitable, cheap and effective artificial feed for fingerlings as well as for adult fish.

While taking note of the inherent handicaps involved in feeding fish, the authors feel that along with the development of a suitable feed, it may be necessary to evolve and improve the feeding technique in order to derive maximum benefits from artificial feeding. Though work on artificial feeding has been carried out by many other workers, emphasis has been on finding out the vitamin, mineral and other food requirements. The present work is mainly concerned with evolving a practical nursery diet for the fry of Indian carps.

2 MATERIAL AND METHODS

The experiments were conducted in 10-litre glass jars with equal quantities of tap water; this eliminated the possibility of introducing extraneous food items. The same number of hatchlings of Cyprinus carpio or Labeo rohita of the same parentage were introduced into each jar. Fry were reared for 15 days, feeding them with uniform quantities by volume of the various feed items, no attempt being made to determine the optimum requirements. The numbers of replicates were increased from two in early screenings to six in later experiments, since less effective feeds were eliminated with the progress of the experiments. Water in all the jars was partially replaced once in three days with fresh tap water, after cleaning and removal of accumulated unused feed and excreta. All fry were fixed in formalin on the sixteenth day for growth and survival rate analysis. Controls without feed yielded the same negative results in the first five sets of experiments, and were therefore not repeated.

Advantage was taken of the experience of earlier workers like Schäperclaus (1933), Alikunhi (1952), Alikunhi et al. (1954; 1955) and many others, who had established zooplankton as the main natural food of carp fry, and hence all items of feed were compared with this standard food. Growth and survival in individual jars were taken as the criteria for deciding the comparative merit of the different items of feed.

Uniformity was maintained in initial feeding (Palmer et al., 1951) and fry were fed from the third day after hatching. Consistency and availability of feed as possible factors affecting growth (Wolf, 1951b) were also taken into consideration. Feeding frequency was limited to once a day, though opinions differ on this issue (Palmer, loc.cit.). While the efficiency of each item was tested independently, no attempt was made to improve this efficiency in combination with any nutritionally inert or binding material.

The more important items including zooplankton were analyzed chemically to determine their protein, carbohydrate, fat and ash contents. Vitamin assay had, however, to be omitted for lack of laboratory facilities.

Selection of feed items was made taking into consideration: (i) economy, (ii) availability, (iii) ease of processing, (iv) keeping quality, (v) non-disintegrating and non-soluble nature, and (vi) consistency. The various items were first sun-dried thoroughly and thereafter powdered by hand or with a mortar and pestle. The powder was sieved through a fine mesh wire netting and kept in air-tight containers.

After the preliminary testing of individual items, poorer feeds were eliminated and the more promising items were tested in combinations of two, three, four and five by repeating the experiments. The particular mixtures were also prepared in different ratios of the constituent items.

The data collected from the fry reared in 195 jars were statistically processed to determine the comparative merit of each item and combination of items.

3 EXPERIMENTS AND OBSERVATIONS

Feeding experiments were carried out both in the laboratory and field during the years 1964 and 1965 with various items of feed singly or in combinations.

3.1 Single items

Nineteen items, viz. jowar (Sorghum vulgare), ragi (Eleusine coracana), Italian millet (Setaria italica), cow pea (Vigna catjang), Bengal gram (Cicer arietinum), horse gram (Dolichos biflorus), maize (Zea mays), cotton-seed (Gossypium herbaceum), duck weed (Lemna minor), mustard oil cake, rice bran, wheat bran, notonectids (Anisops sp.), fish roe, prawn (Macrobrachium lamarrei), fat-free fish protein, fish protein with fat, hydrolyzed protein and poultry feed, were tested. Besides the above items, additions of common salt or yeast in very low percentages were also tried.

Since fry fed with zooplankton yielded the best results in all jars, the optimum growth and survival is based on these results. The average growth and survival were 16.63 mm and 94 percent respectively.

Out of the 19 items given above, the data on 16 items only were studied in detail, deleting rice bran, horse gram and hydrolyzed protein, which yielded very low survival with almost negligible growth. In the remaining observations, since the differences in growth and survival between feeds were quite distinct, further repetition of experiments was not considered necessary. However, in the case of treatments which were repeated, there was no significant difference in either growth or survival. Since hatchlings of uniform size and age were used, growth measurements taken for the analysis were the final measurements. Analysis of variance was performed to see whether the treatments differed significantly.

Table I

Analysis of variance in single item trials

‘F’ is highly significant, which means that the samples differ significantly. But the value of ‘F’ gives only an overall significance and does not indicate the individual differences. Hence, for studying the individual difference between the samples, ‘t’ test was employed. The t value was 1.96 for significance at 0.05 level; i.e. difference between means = 1.96 × SE = 1.96 × 0.1905 = 0.3734 or say 0.4, for significance at 0.05 level. The smallest difference between any two sample means which is significant at 0.05 level is 0.4.

From Table II, it is clear that though the percentage of survival is relatively less in the case of fish roe, due to its comparatively high growth result, it tops in ranking. However, fish roe did not sufficiently satisfy some of the important selection qualities, viz. availability, economy and keeping quality, and had to be dropped.

Further combination experiments were undertaken deleting poultry feed, maize, jowar, cotton seed and Italian millet.

3.2 Two-item mixtures

Out of 25 combinations of two items tested, the results of two combinations, viz. mustard oil cake plus rice bran, and prawn plus fish roe, were found to be very poor. Deleting the above two, the results of 23 combinations were analyzed for ranking. The analysis of variance is as shown in Table III.

‘F’ is highly significant and so the samples differ significantly. By the application of the ‘t’ test, the least difference between sample means which is significant at 0.05 level is found to be about 0.5.

The most striking feature of the results of the experiment with two-item mixtures has been the elimination of the prevailing popular artificial feed, viz. mixture of mustard oil cake with rice bran, as a comparatively inefficient feed. From Table IV it is clear that notonectids plus cow pea in the ratio of 2:3 ranks first whereas the same items in the ratio of 1:1 dropped to the thirteenth rank. In double mixtures, yeast appears nutritionally quite effective.

3.3 Three-item mixtures

Since the efficiency of notonectids, prawn and cow pea (NPC) had become apparent in early trials they were used mainly for the three item combinations. Out of 20 combinations in different ratios tested, data of 17 were studied in detail deleting three, viz. NPC in the ratios of 2:5:3, 7:10:3 and 9:9:2. The analysis of variance for the 17 combinations was as shown in Table V.

Table II

Results of experiments with single item feeds, arranged in the order of efficiency of the feeds

Table III

Analysis of variance of the results of experiments with two-item feed mixtures

Table IV

Results of experiments with two-item feed mixtures arranged in the order of efficiency of the mixtures

Table V

Analysis of variance of the results of experiments with three-item feed mixtures

Table VI

Results of experiments with three-item feed mixtures, arranged in the order of efficiency of the mixtures

Table VII

Analysis of variance of the results of experiments with four-item feed mixtures

‘F’ once again is highly significant. By the ‘t’ test, the smallest difference between sample means which is significant at 0.05 level was found to be 0.5. The results are given in Table VI.

From Table VI, improvement in growth and survival is more obvious in combinations of three items than in two item mixtures. Combinations of NPC in the ratios of 5:3:2 and 3:5:2, rank one and two respectively. It is interesting to note that with 10 combinations of NPC in different ratios, their nutritive behaviour had widely fluctuated and extended to the two extremes in ranking.

3.4 Four-item mixtures

With the indication in improvement of the diet in combination of three items, further attempts were made with mixtures of four items. Six combinations with NPC as the major constituents and others shown in Table VIII as the fourth item, were tested. The analysis of variance for the four-item mixtures is given in Table VII.

Thus ‘F’ remains highly significant with the smallest difference between sample means estimated as 0.4. The results of the analysis are given in Table VIII.

Table VIII

Results of experiments with four-item mixtures, arranged in the order of efficiency of the mixtures

The above suggests no appreciable improvement in growth and survival compared with three-item mixtures.

3.5 Five-item mixtures

With only two combinations tried, by ‘t’ test, significant difference (t = 3.053) was observed. This ‘t’ is with 82 d.f. and as the table value at 0.05 level, viz. 1.989, is much less than the calculated value, the treatments differ significantly.

The results of the two treatments with the average growth and survival data are given in Table IX.

Table IX

Results of experiments with five-item feed mixtures, arranged in the order of efficiency of the mixtures

The above results indicate a negative trend in the effect of the diet.

From the foregoing analyses the following treatments may be selected on the basis of their ranking, excluding fish roe.

Table X

The feed items and mixtures according to their rankings

Table XI

Analysis of variance for the 18 experiments

Here also ‘F’ is highly significant and hence the treatments differ significantly. The smallest difference between sample means which is significant at 0.05 level is 0.5. In Table XII, the results are finally assessed.

Table XII

Feeds and feed mixtures arranged in the order of efficiency, as judged by growth and survival of carp fry

This table shows that a mixture of notonectids, prawn and cow pea in the ratio of 5:3:2 was most effective as a balanced artificial feed for carp fry. It has also been established that notonectids followed by prawn in combination with other feeds invariably yielded satisfactory results, thereby indicating their superiority in nutritive value, as well as their importance as the initial raw material.

4 CHEMICAL ANALYSIS OF IMPORTANT FEEDS

An attempt was made to analyse chemically the important items of feed used (Table XIII). The material was first thoroughly dried and powdered before analysis. Protein level was estimated by Kjeldahl's digestion method and fat was determined by other extraction. Ash was estimated by total ignition and carbohydrate calculated as difference.

Table XIII indicates a high level of protein in notonectids, and carbohydrate in cow pea, while prawn is rich in ash and fat.

Table XIII

Chemical analysis of items of feed

5 FIELD EXPERIMENTS

After the conclusion of the laboratory experiments, the relative merits of the selected diet were tested by conducting limited field experiments.

For this, ponds of comparable size and with more or less uniform ecological conditions were stocked with fry of Catla catla, Labeo rohita, Cirrhina mrigala and Hypophthalmichthys molitrix at the same rate of one million per hectare; here they were reared separately for 15 days. The fry were fed with the notonectids-prawn-cow pea (NPC mixture) diet, keeping mixture of mustard oil cake and rice bran as control feed in duplicate ponds. The feeding frequency was confined to once a day, the feed being broadcast on the water surface. The rate of feeding in each pond was:

On the sixteenth day, fry from all the ponds were netted, their number assessed and growth recorded. The final data are as given in Table XIV.

Table XIV

From the above data, the superiority of NPC mixture over the commonly used mustardoil-cake-rice-bran mixture as a supplementary feed is obvious and could be considered as a suitable nursery diet.

¹ Initial weight of fry stocked was determined on the basis of 0.0014 g as average weight of fry.

6 DISCUSSION

In discussing the comparative merits of the above suggested mixtures as a fish diet, it may be relevant to consider the various feed items that are widely used in other countries as fish feeds. These include beef liver, beef heart, horse meat, hog spleen (frozen), meat scraps, tripe, lard, tankage, liver oils, whitefish meal, canned fish, cooked carp, silkworm pupae, white ants, gelatin, white of egg, dry butter milk, dry skim milk, casein (commercial), rice or wheat powder, rice bran, tapioca, cellu flour, oatmeal, soybean meal and milk, oil cakes, penicillin mat, brewer's yeast, Ipomoea stem, water hyacinth shoots, etc.

From the above incomplete list of supplementary feeds, it is obvious that the protein-rich costly items are beyond the means of fish farmers in a developing country like India. Hora and Pillay (1962) have recorded the various supplementary feeds used by fish culturists in the east and south-eastern countries of Asia. Among them animal material, except silkworm pupae and white ants, is conspicuous by its absence.

As the inherent productive potential of a pond is utilized not only by the fish, but also by the other plants and animals, it may be argued that fish production is likely to be greater if these other organisms are absent, unless they can be used for fish production. The present work shows that considerable quantities of insects and small commercially unimportant prawns growing in a pond provide an effective protein-rich fish-food, and that these organisms are not waste products.

The use of two protein items, namely notonectids and prawn, in the mixture evolved, could be justified by the well-known nutrition policy according to which a diet is likely to be more complete and better balanced in amino acids if it relies on a variety of protein sources rather than on a single source (Almquist, undated). Recent studies on carp nutrition by Hepher and Chervinski (1965) suggest an increase in fish yield when common carp (Cyprinus carpio) are fed on a protein-rich diet, an observation supported by the present work. Though no growth-limiting influence was apparent on the various species of carp fry tested with the present protein-rich diet, the authors do not claim that this suggested new supplementary feed is the optimum balanced diet in all nutritive requirements. Further work towards that end is essential.

An interesting point observed is its suitability for feeding all the cultured Indian carps, common carp and Chinese carps. Among the Indian carps, Catla catla is known for its selective plankton feeding habit. However, laboratory as well as field experiments demonstrated conclusively that this species could feed and grow satisfactorily on NPC mixture.

The economics of the feed depend largely on the availability of notonectids and prawns from the ponds of a fish farm. Normally, the insects could be collected in large quantities from nursery and rearing ponds throughout the year. Sample collections made from three 0.12 hectare ponds at Cuttack weighed 10,7 and 5 kilograms respectively in single hauls. The ratio of wet and dry weight has been found to be 3:1. Similarly, natural occurrence of small prawns, Macrobrachium lamarrei in nursery and rearing waters in considerable quantity is a common feature of many fish farms in India. While the two important items are thus available at almost no cost, the third item, cow pea, is a common cereal of relatively low cost. Where the prawn is not readily available, dry prawn from the market could be used at competitive rates. The field trials however demonstrated that the expenditure involved in feeding could be considered very favourable in relation to the returns from higher production of healthy fry.

Use of aquatic insects as natural food by insectivorous species of fishes, such as largemouth bass and bluegill bream (Howell et al., 1941), and by most of the perches, murrel, minnows, catfish, etc., is an established practice. Natural feeding, to a limited extent, on insects by common carp and some barbels of the carp family are also recorded. Hence, insects and prawns could be considered as suitable feeds for carps in general, provided these items are made available to them in a consumable and digestible condition. This suggests the possibility of utilizing terrestrial insect populations in fish feed.

Throughout the experiments, both in laboratory and field, none of the fry of the various species showed any symptom of dietary disease, which can also be considered in favour of the diet.

7 ACKNOWLEDGEMENTS

The authors are indebted to Dr. B.S. Bhimachar, Director, for critically going through the paper and to Sri K.H. Alikunhi, former Deputy Director, for suggesting this important problem. We are also grateful to Dr. H. Chaudhuri and Dr. M.T. Philipose for their helpful suggestions in carrying out the work and in finalizing the paper. Thanks are also due to Sri P. Datta, for his useful suggestions to improve the statistical part of the paper.

8 REFERENCES

Alikunhi, K.H., 1952 On the food of young carp fry. J.zool.Soc.India, 4(1):77–84

Alikunhi, K.H., 1956 Fish culture in India. Fm.Bull., Delhi, (20):144 p.

Alikunhi, K.H., H. Chaudhuri and V. Ramachandran, 1954 On the role of plankton in the survival and growth of carp fry in nursery ponds. In Symposium on marine and fresh-water plankton in the Indo-Pacific, Bangkok, January 25th and 26th, 1954. Bangkok, IPFC, pp. 99–103

Alikunhi, K.H., 1955 On the mortality of carp fry in nursery pond and the role of plankton in their survival and growth. Indian J.Fish., 2:257–313

Almquist, H.J., d.u. Proteins and amino acids in animal nutrition. New York, U.S. Industrial Chemicals Co., Division of National Distillers Products Corporation, 32 p.

Brookway, D.R., 1953 Fish food pellets show promise. Progr.Fish Cult., 15(2):92–3

Hepher, B. and J. Chervinski, 1965 Studies on carp nutrition - The influence of protein-rich diets on growth. Bamidgeh, 17(2):31–46

Hickling, C.F., 1962 Fish culture. London, Faber and Faber, 295 p.

Hora, S.L. and T.V.R. Pillay, 1962 Handbook on fish culture in the Indo-Pacific Fisheries region. FAO Fish.tech.Pap., (14):204 p.

Howell, H.H., H.S. Swingle and E.V. Smith, 1941 Animal matter preferred by game fish, five years' experiments indicate. Ala.Conserv., 1941:2 p.

Kawamoto, N.Y., 1957 Production during intensive carp culture in Japan. Progr.Fish Cult., 19(1):26–31

Ling, S.W., 1957 Report to the Government of Thailand on the development of inland fisheries. Rep.FAO/ETAP, (653):50

Mitra, G.N. and P. Mohapatra, 1956 On the role of zooplankton in the nutrition of carp fry. Indian J.Fish., 3:299–310

Natarajan, M.V. and A. Srinivasan, 1961 Proximate and mineral composition of fresh-water fishes. Indian J.Fish., 8(2):422–9

Palmer, D.D., et al., 1951 The effect of retardation of the initial feeding on the growth and survival of salmon fingerlings. Progr.Fish Cult., 13(2):55–62

Palmer, D.D., L.A. Ribinson and R.E. Burrows, 1951 Feeding frequency: its role in the rearing of blue black salmon fingerlings in troughs. Progr.Fish Cult., 13(4):205–12

Phillips, A.M., Jr. et al., 1954 Comparison of hatchery diets and natural food. Progr.Fish Cult., 16(4):153–7

Phillips, A.M., Jr., 1952 The effects of supplemental fats in the diet of brook trout. Progr.Fish Cult., 14(1):19–22

Schäperclaus, W. (Transl. F. Hund), 1933 Text-book of pond culture. Fish.Leafl., Wash., (311):260 p.

Seaman, W.R., 1951 Feeding of cooked carp to hatchery trout in Colorado. Progr.Fish Cult., 13(4):227–8

Wolf, L.E., 1951a Comparison of yeast and penicillin mat as supplements to dry meal diets for brown trout. Progr. Fish Cult., 13(3):117–20

Wolf, L.E., 1951b Diet experiment with trout. Progr.Fish Cult., 13(1):17–21

Wolf, L.E., 1951c Experiment with canned fish as food for trout. Progr.Fish Cult., 13(2):95–7