RESTRICTED FI:DP INS/72/003/3 April 1976

A report prepared for the Project on Shrimp
and Milkfish Culture Applied Research and Training

by

M. Pedini
Fishery Biologist
(Associate Expert)

This is one of a series of reports prepared during the course of the UNDP project identified on the title page.

The conclusions and recommendations given in the report are those considered appropriate at the time of its preparation. They may be modified in the light of further knowledge gained at subsequent stages of the project.

The designations employed and the presentation of the material in this document do not imply the expression of any opinion whatsoever on the part of the United Nations or the Food and Agriculture Organization of the United Nations concerning the legal or constitutional status of any country, territory or sea area, or concerning the delimitation of frontiers.

ACKNOWLEDGEMENTS

The Food and Agriculture Organization is greatly indebted to the following individuals who assisted in the implementation of the project by providing information, advice and facilities:

Mr. E. Hamami who assisted in the experiment on food preferences carrying out a considerable amount of work with tests and helping in the pools, to Mr. E. Duursma and Mr. A. Hanafi for the chromatrograms of the diazimon experiment and to the personnel of the chemistry laboratory. Mr. Soesanto of the Directorate General of Fisheries who discussed the programme of work with the Associate Expert was also of great assistance.

FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS
Rome, 1976

Hyperlinks to non-FAO Internet sites do not imply any official endorsement of or responsibility for the opinions, ideas, data or products presented at these locations, or guarantee the validity of the information provided. The sole purpose of links to non-FAO sites is to indicate further information available on related topics.

This electronic document has been scanned using optical character recognition (OCR) software. FAO declines all responsibility for any discrepancies that may exist between the present document and its original printed version.

TABLE OF CONTENTS

1. INTRODUCTION

1.1 Terms of Reference
1.2 Programme of Work

2. EXPERIMENT ON FOOD PREFERENCES OF MILKFISH FRY

2.1 General
2.2 Materials and Methods

2.2.1 Pools
2.2.2 Fry of Milkfish
2.2.3 Chemical Test and Fertilizers

2.3 Results and Conclusions
2.4 Diazinon Experiment

3. MILKFISH POND MANAGEMENT

3.1 Preparation of the Ponds
3.2 Stocking of the Ponds
3.3 Maintenance of the Ponds
3.4 Fertilization schedule of the Ponds
3.5 Harvesting of the Ponds
3.6 Economics of the Pond Operations

Appendix 1 A COMPLEXIMETRIC METHOD FOR THE DETERMINATION OF DISSOLVED OXYGEN IN WATER

LIST OF TABLES

1. Result in Growth, Length (range) and Weight during the Experiment.

2. Weekly Increases in Length and Weight and corresponding Percentages during the Experiment.

3. Analysis of the Gut Contents (Gut Contents are listed in order of Dominance)

4. Schedule in Pond Management

5. Harvest Data of the “E” Series Ponds

6. Cost-benefit Balance of the Milkfish Pond Operation. Cost of in-puts less Labour.

LIST OF FIGURES

1.(a) Plankton Pool

1.(b) Benthic Algal Pool

2. Scheme of the Gill Rakers in Milkfish Fry, with Separation in μ for Specimens 23.1, 36 and 52 mm long.

1. INTRODUCTION

1.1 TERMS OF REFERENCE

The Government of Indonesia, assisted by the United Nations Development Programme and the Food and Agriculture Organization of the United Nations are engaged in a project whose main purpose is to develop improved methods for the expansion of shrimp and fish production on a national scale; to establish data collection programmes and to provide requisite training.

As part of the project operation, FAO assigned Mr. M. Pedini, a Fishery Biologist from April 1975 for four months to assist Mr. P. Padlan, the milkfish expert of the project, and with the following terms of reference:

1. assist the project in investigations relating to the improvement of milkfish culture

2. participate in the organization of extension workers training

3. assist in the location of seed collection centres and the production and distribution of seed.

1.2 PROGRAMME OF WORK

The plan of work included an experiment on food preferences of the young fry of milkfish Chanos chanos Forskal, to determine optimal nutrition during the early stages of growth with natural feeds. Parallel to this experiment, the Associate Expert was to assist Mr. P. Padlan in the day to day management of the milkfish ponds, in order to get himself acquainted with the techniques and methods of milkfish pond management.

2. EXPERIMENT ON FOOD PREFERENCES OF MILKFISH FRY

2.1 GENERAL

The feeding of the early fry of Chanos chanos Forskal has for a long time been a matter for clarification. However, no complete studies on this subject appear in the literature. This is perhaps due to the fact that the fry stages of this fish are very resistant and do not create serious problems when fed on different diets, and that the investigation was centred mostly on pond management. The determination of a preferred feedstuff, however, can be of importance when a stage of intensive culture is reached, in which a good growth, high stocking density milkfish and cheap food are required.

Even in the case of the adult fishes, it is not yet clear whether the species is a plankton feeder or a benthos feeder, and milkfish culture (based in providing benthic or planktonic organisms) is currently practised although the former on a larger scale. However, production figures per hectare are higher in the latter, reaching 4 000 kg/ha/year in freshwater ponds fertilized with sewage waters. For the brackishwater system, which normally provides benthic foodstuff, figures of 2 500 kg/ha/year are claimed with the use of sophisticated techniques, including selective harvesting (Tang, 1972).¹

It was decided therefore, to start a preliminary experiment on food preferences in order to determine the best growth of the fry up to the pre-fingerling size, and in a second stage of the fingerlings, when offered diets based on plankton and benthos. The experiment would consist in feeding the fry with different feeds in every pool, always in excess of their needs so that lack of food could not be a limiting factor for the growth.

¹ Tang, Y.A. 1972 “Stock Manipulation of Coastal fish farms” in Coastal Aquaculture in the Indo-Pacific Region by T.V.P. Pillay (ed) Fishing News (Books) Ltd. London, 497 pp.

The experiment would be conducted at a laboratory scale in small pools and then replicated in larger containers. This choice was made in view of the availability of only small pools at the beginning of the experiment. The duration was established at four weeks.

2.2 MATERIALS AND METHODS

2.2.1 Pools

Four plastic pools of 87 cm diameter, supported by an aluminium frame were used in this experiment. Two pools were utilised for plankton growth keeping an average depth of 45 cm, and the other two were used for benthic algal growth.

In order to expose the benthic algae to a maximum amount of light the bottom of these pools was raised by placing 20 cm of sand under the plastic sheet. Maximum water depth was thus reduced to a mere 38 cm, considered within the range for the growth of benthic algae, (Fig. 1(a), (b).

Plankton pools (named SP1 and SP2) were filled with green water and aeration was put in to keep a turbulent circulation of water.

In SP1, 16 plastic strips measuring 55 × 10 cm (17 600 cm²) the surface available for the growth of epiphyton. The disposition of these plastic strips was initially at random and later they were placed in a circular frame, concentric, to the diameter of the pool. The frame measured 50 cm diameter from which only 13 strips could be suspended, resulting in extra 14 300 cm² surface only. This was done in order to see a possible preference of the fry for the epiphyton growing in the walls of the pool by comparing the gut contents of SP1 and SP2.

SP3 was filled with 15 cm of soft soil (moisture content about 60 %). This soil was taken from one of the ponds of the project to reproduce as much as possible the condition of the area. SP4 was filled with a 10 cm layer of soil from the same pond, but that previously had been dried to a moisture content of about 20–25 %. This followed the Taiwanese-Philippine techniques of periodical drying of the ponds, and would point out the possible differences in growth with our ponds.

2.2.2 Fry of milkfish

Fry of milkfish from Lassen (Java) were employed in this experiment. The fry were quite young and not yet entirely pigmented. Initially 80 fry were stocked per pool giving a stocking density of 134.6 fish/m². The average weight and length of the fish at stocking were 11.36 mg and 13.34 mm respectively (range of length 12.2–14.8 mm). After stocking, samples were taken weekly to measure growth both in length and weight as well as well as to dissect them and study the gut contents.

The number of samples taken from each pool were 15 the first week (18.75 % of the population considering no mortality), 12 the second week, (18.4 % of the population), 10 the third week (18.8 % of the population) and the rest of the original population the fourth week, in order to calculate the mortality during the whole experiment.

Therefore, the stocking densities were decreasing from an initial 134.6 fish/m² to 109.4 fish/m², 89.2 fish/m² in the first, second, and third week respectively.

2.2.3 Chemical test and fertilizers

A series of routine tests was run throughout the experiment with the purpose of controlling water conditions and to determine the adequate time and amount of fertilizers to be used.

Temperature and salinity were checked on a daily basis, and oxygen, every two days in the mornings.

For oxygen determination a complexometric titration method was employed instead of the commonly used Winkler. This complexometric titration method (Appendix 1) modified by Roskam and de Lange (inedit), is based on the reaction of O₂ with ferrous iron to form ferric iron, which is titrated with EDTA in the presence of salicylic acid as indicator. The method offers several advantages over the Winkler for field use, and is strongly recommended for its simplicity, rapidity and stability of the reagents used, which make unnecessary the frequent periodical checking as is in the case of the Winkler. Accuracy is lower than with the Winkler method (0.1 against 0.01) but is sufficient for field use. Besides, possibility of mistakes by unskilled laboratory personnel is by far lower than in the Winkler.

Other tests, performed on a routine basis were pH (every three days or more often if required), and dissolved P₂ in the water, although the results of the latter were not entirely reliable. This was caused by the lack of experience in this particular test with the reagents used, and by wrong calibration curves used initially. Only after two weeks of practice were the results meaningful.

Some alkalinity tests were attempted (twice weekly) but still the laboratory experienced troubles obtaining extremely low values, which would not be realistic.

Unfortunately it was not possible to conduct tests on N₂ compounds with the available materials at the chemistry laboratory. It was the same for any kind of test on nutrients in the soils.

Pools SP1 and SP2 were initially fertilized with 6 ppm n₂ and 1 ppm P₂ using (NH₄) SO₄ and TSP as sources, to produce a plankton bloom.

SP₃ and SP₄ were not fertilised initially since the soil coming from the ponds was supposed to be already rich in nutrients. However, the impossibility of quantifying this amount of nutrients in the soils by means of chemical tests, has been a serious hindrance in this experiment.

Later, at intervals, fertilization of the pools was carried out to replace nutrients consumed by the algal populations, based on the control tests carried out periodically.

2.3 RESULTS AND CONCLUSIONS

Due to the poor development in the environmental conditions of some of the pools during the experiment, the results cannot be considered as entirely reliable for a definite statement The excessive fluctuations in the algal populations due to the rapid blooms and following depletion of nutrients in SP1 and the heavy infestation with Chironomid larvae occurring in SP3 and SP4, together with the small size of the pools used, have been responsible for it.

Only after the completion of the replicates in bigger pools, which will ensure a better control of the fluctuations in the algae, can a definite statement be made.

However, some useful indications can be extracted from this preliminary experiment, based on the study of the algae present in the pools, the differences observed in the four pools during the experiment with respect to growth in length and weight, and the analysis of the guts' contents.

In SP1 and SP2 Chlorophyceae were dominant, more than 90 % in both cases, the average size of the cells 3.5 μ in diameter. The lack of taxonomic literature and classification keys, prevented the correct classification at species level, but they could be identified as Chlorella spp. The remaining 10 % were mainly flagellates of a maximum size of 25 μ and an average size of 12 μ which could be tentatively classified as Symbiodinium microadriaticum (Freudenthal) and Melosira spp (Bacillariophyceae) in SP1. In the last days of the experiment, these pools developed Enteromorpha spp attached to the walls.

In SP3 the dominant bottom population were diatoms, Navicula spp and Amphora spp, the last ten days of the experiment. About ten days after the start of the experiment, a thin film of Chlorophyceae developed only in the surface of SP3 and SP4 but there was no growth of other planktonic algae.

The genus of benthic algae found in SP3 in order of dominance was Navicula, Amphora, Nitzschia, Gyrosigma, Donkinia and Cocconeis.

Algal growth was also observed on the walls of the pool and samples taken from it showed a great dominance of Amphora spp on which the fish from this pool were feeding intensively during the last two weeks. During the last week Enteromorpha started to develop in the pool.

SP4 contained mostly blue-green algae during the experiment. At the beginning two different species of Oscillatoria were identified as dominant. Also a number of Lyngbya spp were observed in the first days. Ten days after the beginning of the experiment Chlorophyceae developed in the surface and at the same time the blue-green algas started to deteriorate as a result of the heavy infestation with Chironomid larvae.

Fertilization did not produce the expected recovery and therefore diazinon was applied to destroy the Chironomids. The results of this application are reported in paragraph 2.4.

Three days after application of diazinon, neither photosynthesis nor activity were observed in SP4 algae, and to be able to continue the experiment it was decided to inoculate the pool with Oscillatoria from the ponds. This was done in order to allow the fish to grow with an amount of food which would not limit the growth as mentioned in 2.1 and to try to restart the growth of Oscillatoria in the bottom of the pool.

The second purpose was only partially achieved and three more applications were necessary to maintain an adequate supply of food. Some Gyrosigma spp were noticed in the bottom samples of SP4, probably introduced with the applications.

The second purpose was only partially achieved and three more applications were necessary to maintain an adequate supply of food. Some Gyrosigma spp were noticed in the bottom samples of SP4, probably introduced with the applications.

The differences observed in the growth of the fry are indicated in Table 1. Mortality rates, for the four pools were 0 % in SP 1, 2.5 % in SP2, 0 % in SP3 and 3.7 % in SP4.

The absolute results show a very clear superiority of the fish placed in SP3 and SP4. However, the results obtained in SP4 with supplementary input of Oscillatoria spp are by far the best, although they must be considered biased with respect to SP3 where no diatoms were inoculated to compensate the loss caused by the application of the diazinon.

They can, however, be directly compared with those obtained in SP2 where planktonic algae were always in excess for the fry.

It is also worth while to examine the differences in the weekly increases and the respective percentages, which are shown in Table 2.

It will be noticed that the percentages of the increases in body length and weight were always inferior in the plankton pools if we discard the results of the second week in SP3 and SP4, in which the infestation with Chironomid larvae and the subsequent trial with the diazinon altered the growth of the fish. The conditions of existence in the pools, as per the tests conducted periodically appear comparable.

It can be seen that the growth of the fry decreased in the plankton pools, stopping practically despite the fact that during the first two weeks the growth had been acceptable, and that algae were always abundant.

This happened when the fish reached an average length of 24 mm and can be related to the size of the available algae and the separation of the gill rakers as well as to a change of behaviour, the fish becoming bottom feeders.

The rate of growth in SP4 was also remarkable. Since apart from the second week they were always offered food in excess the values ranging between 43 and 46 percent (increase in length) could be considered as nearly optimal for this fish in the early stages of growth.

The analysis of the guts' contents was performed as soon as possible after sampling of the fry, in order to avoid post mortem digestion by bacteria present in the digestive tract, which were not fixed by the formaline. The results of the analysis are indicated in Table 3.

From the results obtained in this preliminary experiment some indications can be extracted. First of all it appears that milkfish fry will feed on a wide variety of organisms, both vegetal and animal as demonstrated by the numerous Chironomid larvae, Nematods and Copepods found in the guts' contents. This fish can be considered as an opportunistic feeder with clear herbivorous preferences. Secondly, after an initial period of two weeks, the growth is better on benthic algae and it also appears that the fish undergo a change of behaviour becoming more inclined to bottom feed. This is shown in the change in composition of the guts' contents in SP2 parallel to a slow-down in growth due to the lack of benthic foodstuff in this pool, and despite the abundance of phytoplankton.

Another trend that appears in this preliminary experiment is the preference of the fish for food of a filamentous shape, perhaps due to its biting pattern and form of the mouth. The fact that the guts' contents of SP4 specimens in the first week contained undigested Oscillatoria, can be explained by the insufficient musculature of the gizzard. It can be seen that when the fish are longer than 25 mm, the Oscillatoria contained in the gizzard are crushed, thus allowing a better digestion of the cells and that at lengths inferior to 25 mm the fish is unable to crush the algae.

As a final consideration it should be pointed out that the gill rakers separation in the fry range from 10 μ (see Fig.2) for a specimen of 23.1 mm, to 12 μ for 36 mm length, and 20 μ for a 52 mm fish. A separation of more than 10 μ would be considered excessive for the filtering of the smallest phytoplankton, especially if related to a high swimming speed as it is in the case of milkfish.

Therefore it can be assumed that after an initial period of three weeks this fish will show a preference for benthic algae of filamentous shape.

In order to ascertain this hypothesis and confirm it on a more reliable basis, replicates of the experiment not only will be run in larger pools but also for a double period, i.e. two months time. It is also planned to try parallel experiments for confirmation of the effect of diazinon on benthic algae.

2.4 DIAZINON EXPERIMENT

Due to the consumption of benthic algae by Chironomid larvae, which were competing seriously with milkfish fry for food in SP3 and SP4, it was decided to experiment on the use of diazinon for their eradication.

Basudin, containing 10 percent of the active product was applied. According to Tang and Chen (1959), ¹ diazinin at a concentration of 0.08 ppm active ingredient would be lethal to the Chironomids in 96 hours.

Water was reduced in SP3 to 25 cm and in SP4, the most seriously affected pool, to 10 cm. There was no record of use of this compound in ponds with so small fry, which at the moment of poisoning the water measured on average 19.6 mm in SP3 and 19.5 mm in SP4. Therefore, to avoid any harm to them, fry were removed from the pools leaving only five specimens as control for 24 hours.

Samples of the water and soil (top layer of about 2 mm) were taken prior to poisoning for analysis by means of gas chromatography. Sampling was repeated 24 hours and 96 hours later, including samples of dying Chironomids.

Salinity at the time of poisoning was 32 ppt and 32.2 ppt for SP3 and SP4 respectively. The results of the test conducted by Messrs. E. Duursma and A. Hanafi were as follows:

In the water and sediment chromatograms, peaks were found after the one corresponding to diazinon, which would be due to natural organic phosphate compounds.

Data obtained from sampling after 24 hours including sediment concentration from the first 2 mm of soil layer gives:

The discrepancy in SP4 (hard sediment) could be due to an uneven distribution of the product's granules, and to sampling.

The fry left as control were not harmed and showed no unusual behaviour. Therefore the rest of the fish of the pools were restocked 24 hours after poisoning.

Chironomid larvae were still observed alive after 96 hours although in very limited numbers.

Both the pools were washed at the end of the period to remove the product. Samples were taken of the benthic algae and heavy mortality was noticed amongst diatomes, (Navicula spp dominant in SP3 and bluegreen algae Oscillatoria spp in SP4, which showed total loss of activity).

This fact suggests a possible toxic effect of Basudin on benthic algae, which will be further investigated in the repetitions of the experiment.

¹ Tang, Y.A., and T.P. 1959 Chen “Control of Chironomid larvae in milkfish ponds”. JCRR Fisheries Series No. 4, 36 pp.

Another interesting fact pointed out by the experiment is that 0.08 ppm of diazinon, though harmless to the fry, will also not be lethal in 96 hours to the whole population of the Chironomid larvae. As a result, further immediate application of Basudin in ponds with the purpose of controlling Chironomid larvae must be regarded as potentially harmful for the normal growth of benthic algae.

3. MILKFISH POND MANAGEMENT

During his stay in the project, the associate expert assisted Mr. P. Padlan in the day to day management of the milkfish ponds.

He was able to participate in the preparation of the ponds prior to stocking, their stocking, maintenance and harvest. The ponds used during the associate expert stage were the E series production ponds and the fingerling ponds F2, B2, B3, and B8.

The original scope in the operation of this production ponds was to dry them and expose the bottom to the air, and then fertilize in order to obtain the maximum production possible. The inputs would be the same in the four ponds in order to have four replicas of the cycle.

It should be pointed out in the first place that the ponds of the project area are not the optimum for milkfish culture. The range of tides is very narrow, the highest tide not passing 0.96 m and as a consequence of this the ponds are quite shallow. In fact it was not possible to keep more than 35 cm water in the ponds during this three months period. Besides that, the clay content in the soil is too high with the result that when dried the dikes shrink considerably provoking leakages which have not been possible to control efficiently up to now.

As a consequence of the water regime of the ponds, it has not been possible to drain and dry the ponds in order to allow a better mineralization of the nutrients and an adequate preparation of the bottoms of the ponds. It is envisaged to use pumps in the future to solve this problem.

3.1 PREPARATION OF THE PONDS

The preparation of the ponds started in April and consisted in: a) construction and setting up of the inner wooden gates in the catching ponds, b) repairing of the dikes facing the main canal, c) poisoning of the ponds in order to avoid unwanted species.

The measures of the wooden gates installed in the ponds are 3.0 m length, 1.0 width, and 1.5 m height. Their measures were adequate for these ponds, ranging between 1.0 and 1.4 hectares. The material used was red mahogany available locally that was painted in order to increase its duration. They were first coated with a layer of creosote and then with one layer of coal tar.

The gate was installed with the help of a flat boat from which it was lowered into the water, then centred with respect to the main gate and set on the bottom. The floor was put in place after sealing both ends of the gate with wooden slabs and soil and exposing the bottom by bailing the water out. The gate had removable screens at both ends and to be made leakproof had a central pair of slabs which could be packed with soil in between. The gate could also be used as a fishing box for the shrimps by removing one slab during low tide, inducing a current against the screen.

The life span of such a gate ranges between five and ten years depending on the cleaning of barnacles.

As a result of the soil's composition in the ponds, with excess of clay, and of the activity of burrowing animals such as crabs and eels, the dikes of the ponds show very often serious leakages. In order to avoid them as much as possible, several trenches were dug out in the most severely affected places, and filled with new dry soil. However, this procedure was not satisfactory and leaks reappeared in a short period of time.

Another problem which it is necessary to face before starting operation in the ponds is the presence of unwanted species of fish, that compete with milkfish for food and that can, in other cases, predate upon them.

In order to eradicate them, Thiodan was used at a concentration of 0.1 ppm. Prior to the application of Thiodan it is necessary to place a fine mesh attached to the bamboo screen in its outer part. The scope of this is to avoid the entrance of small fry into the ponds during the period of application of the pesticide. At the concentration employed, it was seen that after four or five hours, the unwanted fish present in the ponds were dying, and then they were caught with scoop nets.

After this, it is necessary to wash the ponds before stocking with milkfish fingerlings, and to conduct a toxicity test to avoid a fish kill immediately after stocking. To perform this test, a hapa is placed some 2 m from the sides of the pond and about five or six fingerlings, 8 to 10 cm long are placed into it. If after two days there is no mortality in the hapa then the pond is safe and it can be stocked. It is, however, absolutely necessary to avoid excessive handling of the fingerlings placed in the hapa to prevent loss of scales or mucus.

For the application of Thiodan and also of the fertilizers, a Philippine type flat boat was employed. This flat boat is extremely helpful not only for this purpose but especially for soil transportation during construction of the ponds. The application of Thiodan and the fertilizers required two men in the flat boat, one of them broadcasting the products. Using this system, the products are distributed more evenly and in a shorter time.

3.2 STOCKING OF THE PONDS

The four E ponds were stocked with fingerlings kept in the ponds F2, B2, B3, and B8. The fingerlings were caught by netting the ponds. Random samples were anesthetized with ether, then measured in length and weight. The fish were transported to their respective ponds in plastic bags partially filled with water.

The average weights at stocking were 49.5 – 55.7 g for fingerlings from F2, 16.92 g for B2, 8.55 g for B3 and 24.5 g for B8.

The fingerlings were pro-rated amongst the four ponds (see Table 4) and the final stocking rate was 2 000 fish/ha equivalent to 53 kg/ha.

The programme was to keep the fish in the ponds for a period of two months and then harvest. The average weight expected at the end of this growing period was about 200 g.

3.3 MAINTENANCE OF THE PONDS

The weather during this period was very irregular and although the wet season usually finishes by the end of April, this year it was unusually long with heavy rains that disturbed the growth of the algae for the whole month of May. The consequences of these rains were negative for the ponds since salinities were very low for the season, in the order of 10–17 ppt. As a result of this, the growth of benthic algae, mainly Oscillatoria in the ponds, was severely affected and in the ponds E3 and E4 filamentous green algae of the genus Enteromorpha took over.

Pond E3 was drained by pumping and exposed to the air in April, and then heavily fertilized with 960 kg of rice bran. The purpose of this was to let the soil dry and crack, and to enrich it as much as possible with an organic fertilizer. Growth of benthic algae in E3 did not progress as expected. Unexpected rains flooded the pond reducing the salinity to a very low value after which Enteromorpha appeared in the pond. It was decided to grow the fish on Enteromorpha and to maintain this food by periodic fertilization. Only urea and TSP were employed.

In E4 heavy infestation with Chironomid larvae occurred at the beginning of May, preventing the growth of benthic algae. To eradicate the larvae, the pond was poisoned with BHC at the rate of 0.08 ppm active ingredient. The pesticide affected also the shrimps which were dying as well as the Chironomid two days after poisoning. The fingerlings of milkfish were not affected by the BHC. Following poisoning of E4 the rains hit again the area, further lowering salinities and causing the appearance of Enteromorpha also in E4.

By 18 May and after a few sunny days it rained again causing a sudden drop in salinity. This time E1 and especially E2 were the ponds more severely affected. In E2, which had a better growth of benthic algae the sudden drop in salinity caused a heavy mortality amongst the algae which decomposing poisoned the water by producing H₂S, noticeable by smell, and other toxic substances. The water in E2 changed the colour to a milky white.

In order to avoid a massive fish kill in this pond due to oxygen depletion and accumulation of toxic substances, the pond was limed with hydrated lime at a rate of 100 ppm after which the condition of the pond improved little by little.

In E1 the rains destroyed the benthic algae present but without a heavy decomposition as in E2, and resulted in a plankton bloom. From this time until the end of the experiment, plankton was dominant in this pond although never in big quantities.

3.4 FERTILIZATION SCHEDULE OF THE PONDS

As mentioned in Section 3 the original purpose of this cycle was to obtain the maximum production from the ponds by drying and exposing the bottom of the ponds, after which a heavy fertilization had to be applied at the same rate in all the four ponds.

Fertilization started with 960 kg/(782.3 kg/ha) of rice bran in E3 but then the rains prevented the continuation of this schedule. It was then decided to continue the cycle on whatever kind of food would appear in the ponds. After the application of Thiodan, the ponds were fertilized at the rates of 200 kg/ha of rice bran, 45 kg/ha urea and 45 kg/ha TSP.

After stocking, fertilization of the ponds was carried out at weekly intervals with the purpose of keeping the growth of the algae. At the end of the cycle the quantities applied in the four ponds were 1 000 kg/ha of rice bran, 102.7 kg/ha of urea and 120 kg/ha of TSP. To achieve this, no rice bran was added to E3 after the 17 May 1975. The quantities employed in every fertilization can be seen in Table 4.

It can also be seen in the same table that from 7 June 1975 the inputs of nitrogen were reduced and the phosphorous increased. The purpose of this change was to accelerate maturation of the algae and reduce the growth. Mature algae would also grow more epiphyton and accelerate decay, thus being more valuable as food for the fish.

3.5 HARVESTING OF THE PONDS

The ponds were harvested at the end of the sixty days period, and two different techniques were employed. The first was to reduce the water level in the pond during low tide. Once in the catching pond, they were caught in approximately ten minutes by the use of a small net.

This procedure was also followed using a net to push the fish into the central canal of the pond and then into the catching pond by again inducing current during high tide.

In order to harvest totally the ponds, after this, the remnant fish were harvested employing derris root. This was done also as preparation of the pond for the next cycle, eradicating also unwanted species.

Shrimp were also harvested in the four ponds for a total of 248 kg (48.99 kg/ha), although it must be noted that after the application of BHC the harvest of shrimp in E4 was almost nonexistant (0.5 kg). The results of the total harvest as well as the survival rates are shown in Table 5.

3.6 ECONOMICS OF THE POND OPERATIONS

Only the operational cost will be considered in the final balance of the ponds. The revenue from the 1 115.92 kg of milkfish harvested and from the 248 kg of shrimp are considered in Table 6.

It must be noted that production in E1 is probably biased by a probable loss of fish. The low survival or recovery rate in this pond cannot otherwise be explained. Therefore for the final consideration, the results obtained in this pond will be disregarded.

In the computation of the costs the items included are: cost of fingerlings, cost of fertilizers, and cost of pesticides and lime. Labour costs are not considered since the ponds in the area are normally run on a familiar basis and only during harvesting time is labour hired.

From the results obtained and from the operational costs it appears that although the highest production obtained is almost 73 percent of the national gross production on average, per year, this is achieved at an excessive cost, which makes the operation uneconomical.

The purpose of this cycle, however, was to investigate the maximum possible production in the ponds during the dry season on a planned input plus protection of the algal population. It could be considered that in the present conditions of the ponds and under similar situations existing during the cycle, it is likely that a maximum production of about 275 kg/ha/ 60 days growing period, will not be surpassed, unless supplementary feeding is provided.

From this, it results that in order to get some benefit from the ponds, the amount of fertilizers employed must be reduced, although this will probably reduce the production of the ponds.

Another possibility to optimize the benefits still maintaining a high production in the ponds still exists. This possibility is the draining of the ponds exposing the soils to the air, thus allowing a better mineralization of the nutrient salts present in the soil. In this manner it will be possible to obtain a similar crop of benthic algae as food for the fingerlings, with a reduced amount of inputs, thus making the operation still interesting economically. This was the original aim of this cycle, but the lack of suitable pumps to drain the ponds and the adverse weather conditions obliged the choice of the first method.

Appendix 1

A COMPLEXIMETRIC METHOD FOR THE DETERMINATION
OF DISSOLVED OXYGEN IN WATER

R. Th. Roskam and D. de Langen
The Netherlands Institute for Fisheries Investigations, Ijmuiden (The Nethlands)
(Received June 23 1962)

Of the titrimetric methods available for the determination of dissolved oxygen in water, the method of Winkler is most widely applied ^{2, 3}. For use on board ship or in the field it has, however, several disadvantages such as the corrosiveness of the reagents, the elaborate precautions needed for work with polluted water, and the instability of the thiosulfate titrant. The solutions of ferrous salts and of ascorbic acid used in other methods 4–7 are even more unstable.

In the method proposed here, the reagents used are neither corrosive nor hygroscopic and they may be applied in solid form. EDTA is the titrant, so that frequent restandardizations are unnecessary⁸. The chemical reactions taking place during the determination can be summarized as follows: dissolved oxygen reacts with iron (II) to form iron (III) which is then titrated with EDTA solution in presence of salicylic acid as indicator. The difference between the stability constants of the iron (II) and iron (III) chelates of salicylic acid^{9, 10,} is greater than the corresponding difference between the sulfosalicylic compounds, hence salicylic acid is more suitable for the titration of iron (III) in the presence of iron (II). Other iron indicators such as tiron and eriochrome azurol S are less convenient, because of slow reactions in cold solution⁸.

The rate of the reaction of oxygen with iron (II) is strongly dependent on pH. It was found that under the conditions prevailing during the determination, it is complete within seconds above pH 7 but takes days below pH 3. A pH of 7.5 was chosen; at higher pH iron hydroxides precipitate, and do not dissolve immediately after addition of acid, so that titrations tend to drag. For the same reason - to avoid precipitation - ferrous ethylenediamine sulfate (FES) is preferable to Mohr's salt as the source of iron (II). A pH of 2.4 was chosen for the titration stage, not only to avoid interference by atmospheric oxygen, calcium, magnesium and iron (II), but also because at this pH the violet 1:1 ferric salicylate chelate predominates and a sharper end-point is obtained from violet to the bright yellowgreen of the iron (III) - EDTA chelate. For the pH adjustment, tris (hydroxymethyl) aminomethane (THAM) and inaleic acid were chosen; these do not cause precipitation and are convenient to handle. The buffering capacity of the test solution is considerable and no harm is done if not exactly the proper amounts of THAM and maleic acid are added.

During the reaction of dissolved oxygen with iron (II) the sample should be sealed off from the air; the conventional Winkler technique may be applied. The data presented here were obtained with wide-necked bottles, the necks of which provided enough space to carry out the titration in the bottles themselves. As soon as the glass stopper is removed the liquid must be acidified to stop the reaction with oxygen. Air interference during the admixture of the maleic acid can be avoided in various ways. For field use, 0.3 percent of glyceryl monostearate can be added to the solid maleic acid; this makes it float for a moment before dissolution so that the liquid is sealed off by an acidic layer. For laboratory use the application of an alcoholic instead of an aqueous solution serves the same purpose.

EXPERIMENTAL

Ferrous ethylenediamine sulfate. This must not contain iron (III). (To check this, add to boiling water some sodium salicylate, maleic acid and FES; no violet colour should develop). Preparations containing iron (III) are difficult to purify by recrystallization. Better results can be obtained as follows. Prepare a concentrated solution of the impure salt in slightly acidic water, filter if necessary, and add some salicylate and sufficient EDTA solution just to remove the violet colour. Precipitate with an equal volume of ethanol, filter by suction, wash with ethanol and dry in a stream of cold dry air.

Anal.Chim.Acta 28 (1963) 78–81

Procedure

To a sample of water in a weighed glass-stoppered bottle of volume about 200 ml add in the following order, 2 ml of aqueous 15 % sodium salicylate solution (or 250–300 mg of solid), 150–200 mg of FES and 0.5 ml of aqueous 20 % THAM solution (or 80–120 mg of solid). Immediately close the bottle, without enclosing air bubbles, and mix by thorough shaking. (The intensity of the brown colour of the 1:2 ferric salicylate chelate gives a rough idea of the amount of dissolved oxygen in the water). Open the bottle and immediately add 5 ml of an ethanolic 15 % solution of maleic acid (for field use, use 600–900 mg of a mixture of maleic acid ground with 0.3 % glyceryl monostearate in a mortar). The colour changes to a deep violet. Titrate with 0.02 M EDTA solution (disodium salt) until the last trace of violet has disappeared from the yellow-green colour of the ferric-EDTA chelate.

RESULTS AND DISCUSSION

Under comparable conditions concerning the human and the equipment factors, the precision of the unmodified Winkler procedure (Table I). The sharper colour change of a good starch indicator compared with the more gradual facing of the ferric salicylate colour is probably the cause.

Table I

REPLICATE DETERMINATIONS IN ONE BATCH OF AGED IJMUIDEN TAP WATER
(Ca²⁺ 5.0, Mg²⁺ 1.3, HCO₂^- 5.1 meq/l)

The accuracy of both methods is the same for unpolluted water and no significant differences were obtained. The data given in Table II suggest that in the case of a water containing significant amounts of unknown substances, the compleximetric method should give more accurate results. However, the possibility of a reaction between these unknown substances and free dissolved oxygen taking place during the determination always remains.

Table II

ERRORS CAUSED BY ADDED INTERFERING SUBSTANCES^a

^a The water used was aged Ijmuiden tap water (Ca2 - 5.0, Mg² 1.3, HCO₂ - 5.1 meq/l).
^b Sliced raw potato with water (1:3) was incubated for 4 days at 37°, and was highly decomposed by butyric acid bacteria (Clostridium pectinovorum).
^c Ml of filtrate of broth added per 100 ml of tap water.
^d See test.

As shown in Table II the proposed method is less sensitive to inferfering substances than the unmodified Winkler procedure. Iron (III) interferes, of course, but the presence of iron (III) hydroxide has hardly any effect; it was shown that when a millimolar suspension of iron (III) hydroxide (made by aerating a ferrous bicarbonate solution) was mixed with maleic acid and salicylate as used in the test, only 2 % was dissolved in 1 h. Much depends therefore on the state in which the iron (III) is present; with iron-containing waters it is advisable to carry out two determinations - one with and one without addition of FES and THAM - and to substract the results.

The lack of interference from halogens is apparently caused by substitution reactions in the sodium salicylate, for a strong smell of the iodoform type develops on addition of sodium salicylate to samples containing free chlorine or iodine. The substitution of sea water for fresh water has no effect on the accuracy and precision of the method.

SUMMARY

A new method is proposed for the determination of dissolved oxygen in water. The iron (III) formed from FES at pH 7.5 is titrated with EDTA solution in presence of salicylic acid indicator after adjustment of the pH to about 2.4. The method is slightly less precise than the Winkler method for pure waters but more accurate for polluted waters; it is simple and convenient for field use.

RESUME

Une nouvelle méthode est proposée pour le dosage de l'oxygène dissous dans l'eau. Elle est basée sur l'oxydation du sulfate de fer (II) - éthylènediamine par l'oxygèse dissous; le fer (III) formé est titré par l'EDTA, an présence d'acide salicylique comme indicateur.

ZUSAMMENFASSUNG

Es wird eine einfache Methode beschrieben zur Bestimmung von gelösterm Sauerstoff in Wasser. Sie beruht auf der Oxydation von Eisen- (II) - Äthylendiaminsulfat (FES) durch den gelösten Sauerstoff und Titration des Eisen - (III) mit EDTA und Salicylsäure also Indikator.

REFERENCES

1. W. Winkler, Ber., 21 (1888) 1843.
2. Standard Methods for the Examination of Water, Sewage and Industrial Wastes, Am. Pub. Health Ass. Inc., New York, 1955.
3. Deutsche Einheitsverfahren zur Wasser, Abwasser und Schlamm-Untersuchung. Verlag Chemie, Weinheim/Bergstr., 1960.
4. Klut-Olszewski, Untersuchung des Wassers an Ort und Stelle, 9th ed., Springer Verlag, Berlin, 1945.
5. W. Leithe, Die Chemie, 57 (1944) 74.
6. V.X. Syrokomskii and T.N. Bondareva, Zavodsk Lab., 16 (1950) 1194.
7. L. Erdey and F. Szabadvary, Acta Chim. Acad.Sci.Hung., 4 (1954)325.
8. G. Schwarzenbach, Die Komplexometrische Titration:Die Chemische Analyse, Bd 45, F. Enke Verlag, Stuttgart, 1957.
9. A. Agren, Acta Chem.Scand., 8 (1954) 266, 1959) 49.
10. D.D. Perrin, Nature, 182 (1958) 741.

Determination of dissolved oxygen in water
by compleximetric titration (Roskam and Lange, 1963)
Analytica Chimica Acta

Remarks:

1. The FeEs must not contain Fe^III, and this can be checked by adding some sodium salicylate, maleic acid and FES to boiling water. No violet colour should develop. Preparations containing Fe^III are difficult to purify by recrystallization, but they can be improved as follows. Prepare a concentrated solution of the impure salt in slightly acidic water, filter if necessary, and add some salicylate and sufficient EDTA solution to just remove the violet colour. Precipitate with an equal volume of ethanol, filter by suction, wash with ethanol and dry in a stream of cold dry air.

2. No calculation is necessary when only one titration wide-necked, glass-stoppered bottle is used; then the same bottle is also used for the preparation of the EDTA solution. When A ml of EDTA solution is used for titration, the calculation of the amount of mgO₂/litre is:

   

   In this calculation is the molarity of the EDTA solution where V is the volume of the used glass-stoppered bottle.

3. The method can also be applied for use with different glass-stoppered bottles, each of known volume (closed with stopper). When the molarity of the EDTA is 0.01 the calculating formula is:

   

4. Dissolved oxygen may be expressed in ml O₂/litre. The converting factor for sea water is of 0 ‰ chlorinity 0° C until 20 ‰ chlorinity 20° C (Jacobson et al, 1950).

   In the case of remark C the formula is:

   

5. The authors urge the necessity of using mg O₂/litre instead of ml O₂/litre or percentages of saturation, because only mg O₂ represent the real invariant quantity of oxygen which is available for metabolic processes.

6. When series of determinations are required, the titration can be done later, because the solution is stable after adding of THAM.

References:

Jacobson, J.P., R.J. Robinson and T.G. Thompson, 1950: A review of the determination of dissolved oxygen in sea water by the Winkler method. Publ. Scientifique No. 11, (Ass. d'océan. phys. Union Géodésique et Géophysique Internationale.

Roskam, R. Th. and D. de Langen, 1963: A compleximetric method for the determination of dissolved oxygen in water. Anal. Chim. Acta, 28 78–81.

SIMPLE FIELD SET FOR ROSKAM OXYGEN DETERMINATION
The spoons are coloured from the left to the right: red, white, blue and orange, the same as the marks on the bottles with powder of the chemicals. The sizes of the spoons are adapted to the proper quantities of chemicals which must be added. Titration is done with an ordinary 10 ml graduated pipette.

Table 1

RESULT IN GROWTH, LENGTH (RANGE) AND WEIGHT DURING THE EXPERIMENT

Table 2

WEEKLY INCREASES IN LENGTH AND WEIGHT AND CORRESPONDING PERCENTAGES DURING THE EXPERIMENT. (GROWTH WAS CALCULATED FOR EACH WEEK SEPARATELY. THE FIGURES ARE NOT CUMULATIVE).

Table 3

ANALYSIS OF THE GUT CONTENTS. (GUT CONTENTS ARE LISTED IN ORDER OF DOMINANCE)

(d = digested
nd = not digested
R = remarks

Table 4

SCHEDULE IN POND MANAGEMENT

* on 29.3.75 960 kg rice bran

.Table 5

HARVEST DATA OF THE “E” SERIES PONDS

Table 6

COST-BENEFIT BALANCE OF THE MILKFISH POND OPERATION COST OF INPUTS LESS LABOUR. CONVERSION RATE
1 US $ = 415 rp

Inputs/ha								Income/ha		Balance rp/ha
	Fingerlings (no/rp)	Rice bran (kg/rp)	Urea (kg/rp)	TSP (kg/rp)	Thiodan	Lime (kg/rp)	BHC (kg/rp)	Milkfish (kg/rp)	Shrimp (kg/rp)
Pond E1						-	-			- 39 658
Pond E2							-			+ 23 727.5
Pond E3						-	-			+ 1 481.1
Pond E4						-			-	- 8 179.4

Fig. 1 (a). Plankton pool.	Fig. 1 (b). Benthic algal pool.

Fig. 2 Scheme of the gill rakers in milkfish fry, with separation in μ for specimens 23.1, 36 and 52 mm long.