II/E-12
PRODUCTIONS OF TILAPIA MOSSAMBICA PETERS, PLANKTON AND BENTHOS AS PARAMETERS FOR EVALUATING NITROGEN IN POND FERTILIZERS¹

by

J. S. DENDY, V. VARIKUL², K. SUMAWIDJAJA, and M. POTAROS²
Auburn University Agricultural Experiment Station
Auburn, Alabama, U.S.A.

Abstract

Twelve 0.1 ha earth ponds were used for this study. Four ponds received no fertilization, 0-0-0 (N-P-K), four received non-nitrogenous 0-8-2 (N-P-K) fertilization, and four received nitrogenous 8-8-2 (N-P-K) fertilization, the same fertilizer treatments which had been applied for the previous five years. Rates of stocking for each fertilizer treatment were 4,942; 9,884; 14,826 and 19,768 fish per ha. The duration of the experiment was from July to October 13–14. Increases in weights of fish per kg were determined upon draining of ponds. Unfertilized ponds produced an average of 242.6 kg per ha. The minimum and maximum rates of stocking resulted in production of 225.9 and 284.7 kg per ha, respectively. The two fertilizer treatments resulted in productions that were essentially alike. In fertilized ponds the average productions with increase in rates of stocking were 342.0, 695.3, 760.1, and 836.7 kg per ha, respectively. The study included rates of growth and reproduction of fish; abundance of plankton, including a volumetric analysis; and abundance of benthos. Results indicated increase in production with increase in rates of stocking and from 63 to 199 percent greater production in fertilized ponds than in unfertilized ponds. However, the elimination of nitrogen from fertilizer for six years did not result in a decline in fish production when water temperature remained above 22°C.

LES PRODUCTIONS DE TILAPIA MOSSAMBICA PETERS, DE PLANCTON ET DE BENTHOS UTILISEES COMME PARAMETRES POUR EVALUER L'EFFET DE L'AZOTE DANS LES ENGRAIS EMPLOYES EN ETANG.

Résumé

Douze étangs en terre d'une superficie de 0,1 ha ont été utilisés au cours de ces expériences. Quatre étangs n'ont pas été fertilisés, 0-0-0 (N-P-K); 4 ont été traités avec un engrais non azoté, 0-8-2 (N-P-K); et 4 l'ont été avec des engrais azotés, 8-8-2 (N-P-K); soit les mêmes formules qui avaient été appliquées depuis 5 ans à ces étangs. Dans chaque groupe, quatre taux d'empoissonnement ont été utilisés, à savoir; 4 942, 9 884, 14 826 et 19 768 poissons par hectare. L'expérience a duré du mois de juillet aux 13–14 octobre. L'accroissement en poids des poissons dans chaque étang a été determiné lors de la mise à sec. Les étangs non fertilisés ont donné une production moyenne de 242,6 kg par ha. Les taux de charge minimum et maximum ont donné respectivement une production de 225,9 et de 284,7 kg par ha. Les deux formules d'engrais ont donné des productions sensiblement analogues. Dans les étangs fertilisés, la production moyenne, rapportée aux taux d'empoissonnement a été respectivement de 342,0; 695,3; 760,1; et 836,7 kg par ha. L'étude a également porté sur le taux de croissance et de reproduction des poissons, l'abondance du plancton, y compris son analyse volumétrique, et l'abondance du benthos. Les résultats montrent que la production s'est accrue à mesure qu'ont augmenté les taux d'empoissonnement et que la production des étangs fertilisés a été supérieure de 63 à 199 pour cent à celle des étangs non fertilisés. Toutefois, l'élimination de l'azote dans les engrais pendant six ans n'a pas entraîné de réduction de la production de poissons lorsque la température de l'eau est demeurée supérieure à 22°C.

PRODUCCIONES DE TILAPIA MOSSAMBICA PETERS, PLANCTON Y BENTOS COMO PARAMETROS PARA LA EVALUACION DEL NITROGENO EN LOS FERTILIZANTES PARA ESTANQUES

Extracto

Para este estudio se utilizaron doce estanques de tierra de 0,1 ha. Cuatro de ellos no recibieron fertilización, 0-0-0 (N-P-K), cuatro la recibieron no nitrogenada 0-8-2 (N-P-K) y cuatro la recibieron nitrogenada, 8-8-2 (N-P-K). Estos tratamientos de fertilizantes eran los mismos que los aplicados en los cinco años anteriores. El número de indivíduos con que estaban poblados los estanques para cada tratamiento era de 4.942, 9.884, 14.826 y 19.768 peces por hectárea. La duración del experimento fue de julio a 13–14 de octubre. El aumento de los peces fue determinado por vaciamiento de los estanques. Los no fertilizados produjeron un promedio de 242,6 por hectárea. Las densidades de población mínima y máxima dieron produciones de 225,9 y 284,7 Kg por hectárea respectivamente. Los dos tratamientos con fertilizantes determinaron producciones que fueron fundamentalmente análogas. En los estanques fertilizados la producción media al aumentar las densidades de repoblación fue de 342,2, 695,3, 760,1 y 836,7 Kg por hectárea, respectivamente. El estudio abarca las tasas de crecimiento y reproducción de los peces; la abundancia de plancton, incluído un análisis volumétrico; y la abundancia de bentos. Los resultados señalaron un aumento de la producción al aumentar las densidades de repoblación, y un aumento del 63 al 199 por ciento de la producción en los estanques fertilizados respecto a los no fertilizados. Sin embargo, la eliminación de nitrógeno del fertilizante durante seis años no dió por resultado el descenso de la producción piscícola cuando las temperaturas del agua se mantuvieron por encima de 22°C.

¹ From theses by the last three authors submitted in partial fulfillment of the requirements for M.S. degree at Auburn University, Auburn, Alabama, U.S.A.

² Present address: Fisheries Department, Bangkok, Thailand.

1 INTRODUCTION

1.1 Background information

In 1959 experiments were begun at Auburn University Agricultural Experiment Station, Auburn, Alabama, to determine the effects of omitting nitrogen from the N-P-K (N, P₂O₅, and K₂O) fertilizer commonly used in fish ponds. Twelve 0.1 ha earth ponds, which had received the same fertilizer treatments annually since 1959, were used in this study. Beginning in 1959, four ponds received no fertilizer, designated as the 0-0-0 treatment; four received P₂O₅ and KCl, designated as the 0-8-2 treatment or the non-nitrogenous fertilizer treatment; and four received N, P₂O₅, and KCl, designated as the 8-8-2 treatment or the nitrogenous fertilizer treatment. The potassium was included to avoid a possibility of deficiency of this element, so that the comparison would be the effectiveness of nitrogen in the fertilizer mixture. Through the years since 1959 various fishes and stocking rates have been used to measure effects of the three fertilizer treatments. The bottom of each pond sloped gently from a depth of 0.6 m at one end to a maximum depth of 1.5 m at the other. The bottom soil was sandy clay at the shallow end, and gradually changed to soft mud ooze at the deep end. Each pond could be filled and drained individually. Water entering the ponds passed through monofilament Saran screens to eliminate unwanted fish.

This paper reports the results of three concurrent cooperative studies conducted in 1964 to evaluate the omission of nitrogen from fertilizer when the ponds were stocked with Tilapia mossambica Peters at rates of 4,942; 9,884; 14,826; and 19,768 fish per ha with fish that ranged from 3.9 to 7.7 cm in length.

The ponds were drained and refilled in June 1964. The following amounts (kg/ha) of fertilizer were applied to the ponds on 17 and 23 June, 23 July, 19 August and 15 September 1964:

The fish were stocked on 6 June. The net productions were determined for each pond by subtracting the weight of fish stocked from the total weight obtained and are presented in Table I.

From Table I it can be seen that although the fertilizer treatments were replicated the stocking rates were not replicated within any fertilizer treatment.

Table I

Net total production of fish in kg per ha from 6 July to 13–14 October in ponds receiving different fertilizer treatments and different stocking rates of Tilapia mossambica

1.2 Parameters studied

Since the weight/ha of fish produced is the primary concern in fish culture, the net production in weight per unit area was used as a parameter with which others were compared. Field work was done cooperatively by those responsible for studies of the various parameters.

A study of plankton was conducted by Varikul. It consisted of measuring dry weights of standing crops and of estimating volumes of dominant zooplankton, both in units per litre or sample.

A study of benthos was conducted by Sumawidjaja. It consisted of obtaining dry weights and numbers of dominant organisms from samples.

A study of fish was conducted by Potaros. It consisted of determining the growth and reproductive rates, and food habits of T. mossambica.

2 PLANKTON

2.1 Review of literature

Reports on evaluation of inorganic fertilizers for production of plankton have usually been on “complete” fertilizers, i.e. the inclusion of N, P₂O₅, and K₂O. Swingle and Smith (1939) reported the relative amounts of phosphorus to nitrogen utilized by plankton to range from 3:1 to 6:1. Neess (1949) mentioned the ability of aquatic organisms, such as bacteria and bluegreen algae to fix nitrogen so that it would be usable by phytoplankton, but also suggested that fertilization with phosphate only stimulated growth of rooted aquatic plants instead of phytoplankton. Hepher (1962) reported that addition of nitrogen in fertilizer for production of plankton was of greatest value in the spring, but was of little value in the summer. He postulated that nitrogen fixation was more effective in warm water than in cold waters.

The use of plankton as a source of food for Tilapia mossambica has been reported by various authors (Chen, 1953; Vaas and Sachlan, 1952; Hora and Pillay, 1962; Vaas, 1954; Kelly, 1955; and Prowse, 1961). However, there was a lack of agreement as to the value of various phytoplankton species to tilapia.

2.2 Materials and methods

2.2.1 Sampling

Three locations in each pond were sampled at three week intervals with a Kemmerer water sampler. At the shallow end samples were taken at the surface. At the deep end they were taken at the surface at a depth of 1.2 m. One litre of each sample was concentrated in a Foerst plankton centrifuge. The concentrate was transferred to a pre-weighed crucible, dried in an electric oven at 60°C for 24 to 48 hours, stored in a desiccator for 24 hours, and weighed to obtain dry weight of plankton per litre of sample. The remainder of the sample was measured and then concentrated in the centrifuge and preserved in five percent formalin for identifications and counts of zooplankters.

2.2.2 Identification of zooplankters

Permanent mounts on microscope slides were prepared for every dominant species of zooplankton. Identifications were made by the use of the keys given by Pennak (1953) and Edmondson (1959). Complete counts of the larger animals in each concentrate were made by the use of a stereoscopic microscope. For rotifers, subsamples were used and counts were made by the use of the Sedgwick-Rafter counting cell.

2.2.3 Volumes of zooplankton

Since different groups of animals that make up the zooplankton differ greatly in volume, the numbers of individuals of a species give little information regarding the importance of that kind of animal as food for fish. Therefore, numbers of individuals were converted into estimates of volumes in mm³ per 1 of sample. This was done by constructing clay models of the animals that were in each dominant group of zooplankters. The volume of each model was determined by displacement of water in a graduated cylinder. The length, width, and depth of the model were measured and the volume was estimated by the following formula:

Where V = volume, L = length, W = width, and D = depth

A correction factor for volume of a given model was obtained by dividing the displacement of water by the estimated volume obtained from the formula.

Average lengths, widths, and depths of representative individuals in a group of animals were used for values to apply in the formula. The value obtained, when multiplied by the correction factor, was the estimated volume of the average individual of the group. This volume was multiplied by the number of individuals to give an estimate of the total value for the group.

2.3 Results

2.3.1 Dry weights

The dry weights of plankton from samples obtained at different stations and depths in ponds were not sufficiently different to indicate differences in the distribution of plankton within any pond.

The rates of stocking fish had little effect on the dry weights of plankton in fertilized ponds, whereas in unfertilized ponds the dry weights were greater in the ponds stocked with 19,765 fish per ha than in ponds that received lower stocking rates.

The average of dry weights of plankton per 1 of sample from ponds that received different fertilizer treatments and different rates of stocking are presented graphically in Fig. 1. Only on one occasion, 7 August, did the average dry weight of plankton from an unfertilized pond even slightly exceed that of the fertilized ponds with the same rate of stocking (19,768 fish per ha). Of the 16 sets of comparisons in Fig. 1, in 13 instances the average dry weight from ponds that received nitrogenous fertilizer treatments were greater than the average dry weights from ponds that received non-nitrogenous fertilizer treatments.

In fertilized ponds the dry weights increased between August and 18 September, and declined by 9 October. In unfertilized ponds the dry weights were approximately the same on all sampling dates, with some decline in October.

2.3.2 List of zooplankton

The kinds of zooplankton encountered are listed below:

Fig. 1 Average dry weights of plankton in mg/l of samples obtained from fertilized and unfertilized ponds stocked with different numbers of Tilapia mossambica

2.3.3 Estimated volumes of dominant zooplankton

The dominant groups of zooplankters included the Cyclopidae, Diaptomidae, nauplii, Sididae and Daphnidae, Bosminidae and Rotifera. The estimated volumes of Cyclopidae from fertilized ponds were greater than those from unfertilized ponds. The volumes from ponds that received the non-nitrogenous fertilizer treatments were approximately the same as those from ponds that received nitrogenous fertilizer treatment. The volumes of the Diaptomidae were approximately the same from fertilized and unfertilized ponds. The volumes of Sididae and Daphnidae and of nauplii from fertilized ponds were not uniformly different from those from unfertilized ponds. The volumes of Bosminidae were conspicuously greater from unfertilized ponds than from fertilized ponds. The relative abundance of this group according to the rates of stocking of fish was not consistent. The greatest average volume of Bosminidae per 1 of sample was 0.97 mm³, and was obtained on 18 September from the unfertilized pond with the next to highest rate of stocking. The next greatest volume was 0.51 mm³, and was obtained on 9 October from the unfertilized pond with the highest rate of stocking.

3 BENTHOS

3.1 Review of literature

Use of standing crops of benthos to evaluate effects of fertilization of waters has been reported by several authors; Howell (1942), Patriarche and Ball (1949), Ball (1949), Hayne and Ball (1956), Rabanal (1960), McIntire and Bond (1962), Liang (1964) and others. The standing crop consists of the organisms that are left after predation by fish. For this reason Hayne and Ball (1956) and Dendy (1956) warned against use of standing crop as a parameter of productivity unless the nature of the fish population and environmental factors are taken into consideration.

An adaptation of the multiple-plate sampler (Hester and Dendy, 1962) by supporting it with the plates held vertically and with one edge in contact with the bottom was reported by Liang (1964). An electrical stimulator that causes organisms to move and thus be seen easily was described by Bayless (1961) as an aid in separating benthic organisms from sand and debris.

Liang (1964) found that in winter, following summer fertilization, samples from the ponds that had received no fertilization contained less benthos than did samples from fertilized ponds, but the samples from ponds that had received the non-nitrogenous fertilizer treatments were essentially like those from ponds that had received the nitrogenous fertilizer treatment.

The references used in identifying benthos were by Johannsen (1935; 1936; 1937), Thomsen (1937), Pennak (1953), Edmondson (1959) and Usinger (1956).

3.2 Materials and methods

The Ekman dredge was used for part of the sampling. It sampled an area of 225 cm². Samples were obtained at depths of 0.6 m and 1.2 m. The bottom materials were brought to the laboratory for further processing. Those samples that could not be processed immediately were refrigerated. To separate organisms and light debris from the sand and gravel, water was poured into the sample, mixed vigorously, and then poured through a screen. This was repeated several times until no macroscopic organisms remained in the sand. By the use of the electric stimulator, the organisms were located and removed from the debris. Further examination was made to discover and remove dead individuals.

Two uses were made of multiple-plate samplers. Samplers that were composed of 9 large plates and 8 small plates, thus providing 8 protected spaces in which benthos could develop, were hung in such positions that they were approximately 0.3 m from the surface and 0.3 m from the bottom. The durations of exposure were three weeks for the first three sampling periods and one week for the last period. Before being removed from the water each sampler, with the water immediately around it, was enclosed in a plastic bag to prevent loss of loosely attached organisms. In the laboratory each sampler was dismantled in a pan and the accumulation of benthos and “aufwuchs” debris were scraped from the edges and inner surfaces. The sample was concentrated in a Foerst plankton centrifuge, and preserved in 70 percent alcohol. Later the samples were transferred to a preweighed crucible, dried in an electric oven at 60°C for 48 hours, stored in a desiccator, and weighed to obtain the dry weight of the sample. The other use of multiple-plate samplers was that described by Liang (1964). Each sampler, consisted of two large plates separated by one small plate, provided only one protected space in which benthos could develop. The contact of the bottom with the edge of each sampler allowed non-swimming and burrowing forms to have access to space between the plates. When these samplers were dismantled, the organisms were preserved, identified, and counted. All counts and weights are per m².

3.3 Results

3.3.1 Sampling with Ekman dredge

The kinds of benthic animals found in samples obtained by the use of the Ekman dredge are listed below:

Numerically the oligochaete worms were the dominant organisms in these samples. Because of the relatively large sizes of many individuals, these worms constituted the greatest volume of potential food for fish. However the variation in abundance was so great that no distinct trends related to fertilizer treatments or rates of stocking of fish were evident.

Chironomid larvae were more abundant in Ekman samples from fertilized ponds than in those from unfertilized ponds. The numbers obtained from ponds that had received the nitrogenous fertilizer treatments were somewhat smaller than the numbers from ponds that had received the non-nitrogenous fertilizer treatment.

3.3.2 Sampling with muliple-plate samplers touching the bottom

The kinds of benthic animals found in samples obtained by the use of the multiple-plate samplers that had one edge touching the bottom are listed below:

A comparison of the lists of organisms from the two kinds of samplers reveals that a much larger variety occurred in samples on the multiple-plate samplers than in the Ekman samples. The difference was in part the result of the Ekman sample being “screened” in the process of separating organisms from sand and debris. Some of the smaller individuals probably were lost through the screen. However, the difference in the dominance of groups is noteworthy. On the multiple-plate samplers, the chironomids made up the greatest volume of potential fish food. The damselfly naiads were not numerous, but because of the relatively large size of many of them, this group ranked second. The oligochaetes were numerous on the plates, but were less abundant and were much smaller than most of the individuals in the Ekman samples.

In samples from fertilized ponds, the chironomid larvae were more numerous than in samples from unfertilized ponds. The numbers from ponds that had received nitrogenous fertilizer were approximately the same as those from ponds that had received the non-nitrogenous fertilizer treatment. The different rates of stocking appeared to influence the numbers of chironomid larvae. The samples from ponds that received the higher rates of stocking contained the larger numbers of chironomid larvae than did those from ponds with lower stocking rates. Since animals between the plates on the samplers were not exposed to predation by fish, their relative abundance after a period of exposure is assumed to be a parameter of the productivity, which appeared to increase with the increase in the rate of stocking of fish.

The oligochaetes were more numerous in samples from fertilized ponds than in samples from unfertilized ponds, but the samples from ponds that received the nitrogenous fertilizer treatment were essentially like those from ponds that received the non-nitrogenous fertilizer treatment. The stocking rates did not appear to influence the number of oligochaetes.

3.3.3 Dry weights from multiple-plate samplers not touching bottom

These data are entirely quantitative. No separation of organisms occurred in the processing. The samplers 0.3 m from the surface and 0.3 m from the bottom were located near edges of the ponds. In this location the plates offered attractive places for oviposition by many kinds of animals. Usually the egg masses were on the upper surface of the top plate or the lower surface of the bottom plate, and for this reason were not included in the sample. However, when egg masses occurred between the plates they became part of that sample. Such factors resulted in a large amount of variation in the dry weights of samples.

Statistically there appeared to be no difference between the average dry weights from ponds that received 0-0-0 and the 8-8-2 fertilizer treatments, and no difference between the 8-8-2 and the 0-8-2 treatments. However dry weights of samples from ponds that received 0-8-2 treatment were greater than those from ponds that received 0-0-0 treatment. There was no evidence that the differences in the rates of stocking influence the dry weights.

From the studies on benthos in various kinds of samples the 0-0-0 fertilizer treatment was not as productive as was either the 0-8-2 or the 8-8-2 treatment. The 0-8-2 and 8-8-2 treatments were approximately equal in value. The 0-8-2 appeared to be the best when evaluated by dry weight method.

4 FISH

4.1 Review of literature

Evaluations of the use of nitrogen in fertilizers for production of fish in ponds have not given uniform results. The work of Swingle and Smith (1939) was mentioned in section 2.1. Lawrence (1943) reported best production when fertilizer contained nitrogen along with phosphorus and potash. Surber (1943) indicated that nitrogen played an important role in control of weeds. Hepher (1962) reported that phosphate in fertilizer was responsible for a greater increase in the weight of fish than was nitrogen. Hickling (1962) found that nitrogen in fertilizer was a poorer producer than was phosphate. Rabanal (1960) and Gooch (1962), both reporting on work done in the ponds used in this study, stated that elimination of nitrogen from the “complete” (N-P-K) fertilizer did not result in a decrease in the production of fish. Byrd and Swingle (1964) recommended that nitrogen could be eliminated from fertilizer for old, well-fertilized ponds without reducing fish yields. Tilapia mossambica have been found to feed on a wide variety of material including plankton, decomposing vegetation, filamentous algae, periphyton, benthos, and various supplemental feeds (see references cited in section 2.1). Kelly (1955) observed that in aquaria T. mossambica ceased to feed when the temperature of the water was lowered to approximately 15.5°C.

When eaten by T. mossambica, phytoplankters varied in digestibility. Vaas and Sachlan (1952) reported that Botryococcus and Microcystis were incompletely digested. Prowse (1961) reported that Anabaenopsis spp. and Oedogonium were well digested, very little Spirogyra appeared to be digested, but Microcystis, Oscillatoria and Anabaena spp. did not appear to be digested.

4.2 Materials and methods

Beginning 28 July and at four week intervals thereafter, the fish populations in each pond were sampled with a seine. Each sample was composed of 20 fish, which were weighed to the nearest g and measured to the nearest 2 mm total length. Of these, 10 were preserved in 10 percent formalin. The fish that were kept were replaced by individuals of approximately the same size. When the ponds were drained all fish were collected and separated into originally stocked individuals and their young. The final weight of fish minus the initial weight was considered as the net production. Numbers of young fish were estimated by counting the individuals in weighed samples and computing the total numbers from the weights of young fish. The relative growth rate at the time of the draining of the pond was calculated by the formula:

Gut contents analyses were made on the preserved fish. For freshly eaten food the contents of the anterior 2.5 cm of the gut, including the enlarged anterior end or stomach, was removed. The contents of the posterior 2.5 cm of the intestine were removed to determine the digestibility of various foods. The material obtained from each region was mixed with 10 ml of water in a dish and examined with stereoscopic and compound microscopes. The method of “frequency of occurrence” advocated by Lagler (1956), was used to evaluate food habits. The viability of plankton in faeces from fish was tested by placing faeces from one fish from each pond in 100 ml of sterilized water in a 250 ml flask and exposing these cultures to light from a window facing north. The cultures were kept for four weeks.

4.3 Results

4.3.1 Production

The production of fish in unfertilized ponds appeared to be increased by increase in rates of stocking, but results were somewhat variable. The productions in fertilized ponds increased 145 percent with the increased rates of stocking. However, the productions that resulted from the 0-8-2 fertilized treatment and those from the 8-8-2 fertilizer treatment at a given stocking rate were essentially equal. This lack of importance of added nitrogen as a component of fertilizer for these ponds agrees with the results of Rabanal (1960), Gooch (1962), and Liang (1964). The continued increase in production with increase in rate of stocking (Table II) indicated that the maximum rate of stocking used probably was not the optimum rate for maximum production of fish with the fertilizer treatments that were applied.

4.3.2 Rate of growth

The rate of growth usually decreased as the rate of stocking increased. The relative growth rates (Table II) in ponds that received the 0-0-0 fertilizer treatment were lower than in the ponds that received the 0-8-2 or 8-8-2 fertilizer treatments, but the relative growth rates that resulted from the 0-8-2 and the 8-8-2 treatments were essentially the same.

Table II

Relative growth rates of Tilapia mossambica from 6 July to 13–14 October in ponds that received different fertilizer treatments and different rates of stocking

4.3.3 Reproduction

The production of young fish did not appear to be influenced by the rates of stocking but greater weights of young fish were produced in fertilized ponds than in unfertilized ponds (Table III).

Table III

Production of young Tilapia mossambica in kg per ha from 6 July to 13–14 October in ponds that received different fertilizer treatments and different rates of stocking

4.3.4 Food habits

Fish obtained during July and August contained an abundance of food, but there was a decrease in, or absence of, the food-stuff in digestive systems of fish collected on 16 September and 10 October. This agreed with the observations of Kelly (1955) that T. mossambica ceased to feed when the water temperature was reduced to approximately 15.5°C.

The digestive tract is typical of fish that feed on detritus. There is no true stomach in the species; it is replaced by a bulb-like enlargement of the tract. The intestine is long and much coiled, and its wall is quite thin.

T. mossambica fed chiefly on detrius and phytoplankton. Of the fish examined, 91.9 percent contained phytoplankton and 89 percent contained detritus. Zooplankton served as additional food. Microcrustaceans were present in 62.9 percent of the fish. They were Cladocera of the families Sididae and Daphnidae; Copepoda of the Cyclopidae and Diaptomidae, and nauplii; and Ostracoda. From ponds that received 0-0-0 fertilizer treatment, 44.4 percent of the fish contained crustaceans, whereas from the ponds that received 0-8-2 and the 8-8-2 fertilizer treatments, 75 percent of the fish contained crustaceans. Although Varikul (section 2.3.2) found Cladocera of the family Bosminidae to be most abundant in ponds that received 0-0-0 fertilizer treatment, even in those ponds these animals were less numerous in the fish than were other crustaceans. Rotifers, chiefly Keratella were found in 12.3 percent of the fish examined.

Benthos, chiefly larvae of Chironomidae, was eaten occasionally. Other insects and oligochaete worms were observed less frequently. A few immature mayflies and damselflies were found. Rare items were fish fry, bryozoans and fragments of higher plants.

4.3.5 Digestibility

The volume of detritus was reduced from large in the anterior part of the gut to quite small in the posterior end. This was judged to indicate a degree of digestibility of at least part of the material. Most of the zooplankton and benthos, except the Ostracoda, was well digested.

Different kinds of phytoplankton varied in digestibility, as judged by their appearance under the microscope. Oedogonium, Surirella, Anabaena and Spirulina, appeared to be well digested. Eudorina, Oocystis, Pediastrum, Microasterias, Staurastrum, Microcystis, Oscillatoria and diatoms other than Surirella appeared poorly digested. Closterium appeared to be unaffected by digestion.

4.3.6 Viability

None of the cultures of faeces from fish developed phytoplankton. However, Ostracoda occurred in every culture. Some were active almost immediately after the faeces were put into the water. Others appeared later. The numbers of these animals that passed through the digestive tracts as living adults and as viable eggs were not determined. There was no evidence that ostracods were affected by the digestive system of the fish.

5 DISCUSSION

In unfertilized ponds the rates at which fish were stocked influenced the dry weights of plankton. Although the fish fed on plankton, the harvest of plankton by fish at the highest stocking rate appeared to be more than counterbalanced by the added fertility that resulted from the presence of the fish, as the dry weight of plankton per 1 of sample was greatest in the pond that contained the largest number of fish. By feeding on detritus, the fish may hasten the rate of recirculation of plant nutrients in forms that are usable by plankton.

In fertilized ponds, different rates of stocking did not influence the dry weights of plankton. The dry weights of plankton in samples from the 8-8-2 ponds were usually somewhat greater than in samples from 0-8-2 ponds.

Most of the volumes of various kinds of microcrustaceans were greater in fertilized ponds than in unfertilized ponds. The volumes of Bosminidae were conspicuously greater in unfertilized ponds than in fertilized ponds. However, even in unfertilized ponds these small cladocerans were essentially unused by fish.

It was observed (i) that chironomid larvae were the dominant kind of benthos eaten by the fish and (ii) that there were greater numbers of these larvae per unit area in the protected space between plates of multiple-plate samplers than there were in standing crop samples by the Ekman sampler. These observations indicated that the fish had harvested most of the available larvae. They indicated also that when many fish are present the use of multiple-plate samplers may give a better understanding of the productivity of a body of water than is obtained by the use of the Ekman sampler. In contrast to the dry weights of plankton, the abundance of chironomid larvae appeared to be slightly greater in ponds that received non-nitrogenous fertilization than in ponds that received nitrogenous fertilization.

The rates of growth of fish were influenced greatly by fertilization, but the presence of nitrogen in the fertilizer did not increase weights of fish produced. Increased rates of stocking tilapia in unfertilized ponds gave variable results, but indicated a moderate increase in production with increased numbers of fish. In fertilized ponds increase in rates of stocking, from approximately 5,000 to 20,000 tilapia per ha, increased average net production from 341.9 to 836.7 kg per ha, an increase of 145 percent.

Although there was a considerable amount of variation in the details of the results, the data indicate that ponds that received fertilizer produced much higher yields than did the unfertilized ponds. There was no indication that the nitrogenous fertilizer treatment was superior to the non-nitrogenous fertilizer treatment for production of fish during periods while water temperature remained above 22°C.

6 REFERENCES

Ball, R.C., 1949 Experimental use of fertilizer in the production of fish-food organisms and fish. Tech.Bull.Mich.agric.Exp.Sta., (210):28 p.

Bayless, J.D., 1961 The use of electrical stimuli in live-picking organisms from bottom samples. Proc.stheast.Ass.Game Commrs, 15:286–8

Byrd, I.B., and H.S. Swingle, 1964 Pond fertilization cost can be reduced by 66 percent. Ala.Conserv., 34(3):4–6

Chen, Tung-Pai, 1953 The culture of tilapia in rice paddies in Taiwan. Fish.Ser.Chin.-Amer. Comm.rur.Reconstr., (2):29 p.

Dendy, J.S., 1956 Bottom fauna in ponds with largemouth bass only and with a combination of largemouth bass plus bluegill. Tenn.Acad.Sci., 31(3):198–207

Edmondson, W.T. 1959 (Ed.) Ward and Whipple's fresh-water biology. New York, John Wiley and Sons. Inc., 2nd Ed. 1272 p.

Gooch, B.C., 1962 Preliminary observations on the residual nature of nitrogen in ponds and its significance to fish production. Unpublished Master's thesis. Auburn University, Auburn, Alabama.

Hayne, D.W. and R.C. Ball, 1956 Benthic productivity as influenced by fish predation. Limnol.Oceanogr., 1(3):162–75

Hepher, B., 1962 Ten years research in fish ponds fertilization in Israel. The effect of fertilization on fish yields. Bamidgeh, 14(2):29–38

Hester, F.E. and J.S. Dendy, 1962 A multiple-plate sampler for aquatic macroinvertebrates. Trans.Amer.Fish.Soc., 91(4):420–1

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 region. FAO Fish.Biol.tech.Pap., (14):204 p.

Howell, H.M., 1942 Bottom organisms in fertilized and unfertilized fish ponds in Alabama. Trans.Amer.Fish.Soc., 71:165–79

Johannsen, O.A., 1935 Aquatic Diptera. Part 2. Orthorrhapha, Brachycera and Cyclorrhapha. Mem.Cornell Univ.agric.Exp.Sta., (177):1–62

Johannsen, O.A., 1936 Aquatic Diptera, Part 3. Chironomidae: Subfamilies Tanypodinae, Diamesinae and Orthocladiinae. Mem.Cornell Univ.agric.Exp.Sta., 205:1–84

Johannsen, O.A., 1937 Aquatic Diptera, Part 4. Chironomidae: Subfamily Chironominae. Mem. Cornell Univ.agric.Exp.Sta., (210):1–56

Kelly, H.D., 1955 Preliminary studies on Tilapia mossambica Peters relative to experimental pond culture. Proc.stheast.Ass.Game Commrs, 10:139–49

Lagler, Karl F., 1956 Freshwater fishery biology. Dubuque (Iowa), Wm.C.Brown Co., 421 p.

Lawrence, V.C., 1943 Techniques of fish pond management. Misc.Publ.U.S.Dep.Agric., (528):22 p.

Liang, J. K., 1964 Benthic fauna of fertilized and unfertilized ponds in winter. Unpublished Master's Thesis. Auburn University, Auburn, Alabama

McIntire, C. D. and C. E. Bond, 1962 Effects of artificial fertilization on plankton and benthos abundance in four experimental ponds. Trans.Amer.Fish.Soc., 91(3):303–12

Neess, John C., 1949 Development and status of pond fertilization in central Europe. Trans.Amer.Fish.Soc., 76(1946):335–58

Patriarche, M. H., and R. C. Ball, 1949 An analysis of the bottom fauna production in fertililized and unfertilized ponds and its utilization by young-of-the-year fish. Tech.Bull.Mich.St.Coll.agric.Exp.Sta., (207):35 p.

Pennak, R. W., 1953 Fresh-water invertebrates of the United States. New York, The Ronald Press Co., 769 p.

Prowse, G. A., 1961 The use of fertilizers in fish culture. Proc.Indo-Pacif.Fish.Coun., 9(2):73–5

Rabanal, H. R., 1960 The effect of no fertilization and non-nitrogenous fertilization upon the chemistry of water, the plankton, bottom organisms and fish production in ponds that had received complete (N-P-K) fertilizers for the preceding 15-year period. Unpublished Doctor's Dissertation. Auburn University, Auburn, Alabama.

Surber, E. W., 1943 The effects of various fertilizers on plant growths and their probable influence on the production of smallmouth black bass in hardwater ponds. Trans.Amer.Fish.Soc., 73:377–93

Swingle, H. S. and E. V. Smith, 1939 Fertilizers for increasing the natural food for fish in ponds. Trans.Amer.Fish.Soc., 68(1938):126–35

Thomsen, L. C., 1937 Aquatic Diptera, Part 5. Ceratopogonidae. Mem.Cornell Univ.agric.Exp.Sta., (210):57–80

Usinger, R. L., 1956 Aquatic insects of California. Berkeley and Los Angeles, Univ.of Calif. Press, 508 p.

Vaas, K.F., 1954 On the nutritional relationships between plankton and fish in Indonesian freshwater ponds. Proc.Indo-Pacif.Fish.Coun., 5(2):90–7

Vaas, K.F. and M. Sachlan, 1952 Notes on fisheries exploration of the artificial lake Tjiburuj in West Java. Contr.gen.agric.Res.Sta.Bogor., (128):1–22