Some Observations on Fish Feeding in Integrated Fish Farming Ponds based on Delta C Analyses of Fish Flesh and Natural Foods

by

Shan Jian, Chang Laifa, Gua Xianzhen, Fang Yingxue,
Zhu Yun, Zhou Xiaoxing and Zhou Enhua
Regional Lead Centre in China
Asian-Pacific Regional Research and Training Centre
for Integrated Fish Farming, Wuxi, China

and

Gerald L. Schroeder
Institute of Animal Sciences
Agricultural Research Organization
P.O. Box 1185, East Hampton
N.Y., U.S.A. 11937

NETWORK OF AQUACULTURE CENTRES IN ASIA
Bangkok, Thailand
September 1983

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.

Some Observations on Fish Feeding in Integrated Fish Farming Ponds
Based on Delta C Analyses of Fish Flesh and Natural Foods

INTRODUCTION

Integrated fish farming, an efficient method of combining fisheries, agriculture and livestock to increase the overall production and obtain maximum utilization of resources is a traditional practice in China. High fish production has been achieved. In order to transfer and adapt the Chinese technology in developing countries having different agro-climatic conditions, there is a need to establish through research a better understanding of the biological processes that are taking place in fish ponds under the system of integrated fish farming.

The fish pond is often treated as a “black box”. We stock fish; we add manures and grasses and in some cases chemical fertilizers or processed feeds. Via photosynthesis algae grow. In several months, if we have managed the pond well, marketable fish are harvested. What the actual sources of the fish growth were we can only estimate from analyses of the gut content. But such analyses reveal only the foods most recently consumed and almost always the major component is “unidentified organic matter”.

A tracer is needed that will show incorporation of food into the fish's body. Radioactive tracers (¹⁴C: ³²P) are useable only in very small ponds or tanks. Even then the feeds must be artificially labelled with the radioactive isotope.

There is in nature a naturally occurring tracer that is present in all organic matter. It is the ratio of the 2 stable isotopes of carbon (¹³C: ¹²C). This ratio, reported as delta C, varies among plant species in a way dependent upon the photosynthetic pathway of the particular plant. The delta C of animal bodies is similar to the delta C of the foods they assimilated. Hence if a sufficiently large range of delta C values exists among the available foods, it is often possible to estimate which foods or groups of foods were assimilated by an animal merely by comparing the delta C of that animal with the delta C of the available foods.

Clearly there are limitations to the method. We are using one tracer in a multidimensional system. When several foods are available, it may be possible only to distinguish between 2 groups of foods, each group containing several possible foods with similar delta C values. Even with this limitation the delta C isotope method has given insights into complex food webs.

STABLE ISOTOPES: BACKGROUND

All natural carbon contains 2 stable isotopes: ¹²C (atomic weight 12) and ¹³C (atomic weight 13). ¹²C accounts for approximately 99% of the total weight. ¹³C accounts for approximately 1%.

To avoid cumbersome fractions and to permit interlaboratory comparasions, the ¹³C: ¹²C ratio is usually reported as:

The standard is the PDB carbonate rock of South Carolina, U.S.A.

The delta C of plants varies among species according to the photo-synthetic pathway that the species used to fix CO₂ from the atmosphere. Two general classes of plants exist: C₄ plants with delta C value of -10 to -16 and C₃ plants with delta C values of -22 to -34. In Table 1, examples of C₄ and C₃ plants, are listed. Note that the lipid fraction usually has delta C several o/oo more negative (i.e. more ¹²C and therefore “lighter”) than the non-lipid parts of the plants.

Animals show considerably less fractionation of the carbon they assimilate than do plants. Hence the delta C of an animal is similar to the delta C of the foods it eats. The delta C of lipids of animals are several (about 5) o/oo more negative than the flesh delta C. Depending upon the % fat in the animal, the flesh delta C is usually 1–2 o/oo less negative than the whole body delta C. With fish, a reasonable estimate is that flesh delta C minus 1.5 equals whole body delta C. Hence for large fish, when homogenizing the entire body is not practical, the whole body delta C can be estimated from a sample of the flesh. Fish tail fin delta C has been found to be about 1 o/oo less negative than the flesh. For example:

DELTA C SURVEY

Samples of various food components which contribute to fish growth in ponds of an integrated farm were collected for delta C analysis. Also sampled were fish flesh and fin, pig manure and vegetation. Sampling procedures are briefly described.

Fish flesh and fin

A one-mu* pond and a 5-mu, 1.5 m and 3 m deep respectively, were stocked with fish on 15 April 1983. On 26 July samples of the fish were netted, weighed and portions cut and dried for delta C analyses. Details of stocking and sampling are listed in Table 2.

* 1 nu = 667 m²

Due to the numbers of fish species being sampled, homogenization of whole fish was not practical. On the side of each fish, an area of approximately 2 cm square was scraped free of scales. With a scalpal, the skin was cut back from 1 cm square, and a piece of flesh 1 cm square × 0.5 cm thick cut out. In the bighead carp the procedure was modified. A portion of tail fin was cut, the size being enough to provide 10–20 mg dry weight. The samples were rinsed in tap water, drained momentarily and dried at 90°C overnight in open, labelled petri dishes.

Natural foods and supplied manures

On 1 August, the natural foods and supplied manures of the 1-mu pond were sampled. A cylindrical water sampler, able to contain 1 liter was tossed into the pond at approximately 10 locations around the pond, 1–2 m from the bank. Filtering another water sample of pond water through a 37 micron net (400 meshes) showed no visible retained matter in 10 liters of water. It was decided not to filter the water from the sampler. The 10 samples were mixed in a bucket and 1 liter taken to the laboratory where it was centrifuged. Prior to drying the sample was observed by microscope and composed of mostly algae with some protozoaes. The precipitation was dried at 90°C.

Pond bottom sediments

Surface sediments from the pond bottom were sucked into a remote sampler held approximately 1.5 m from the pond bank at 10 locations around the pond. One sample, snelling distinctly of pig manure was discarded since pig manure was to be analysed separately. The samples were mixed and observed under 100X and 400X microscope. Some algae and protozoans were present. Half the sample, still moist, was acidified using concentrated H₃PO₄. Colour paper indicated pH 1 or 2. Slight to moderate effervescence observed as CO₂ from the carbonate was liberated. Acidified and not acidified sediments were dried at 90°C. A portion of each sediment was analysed for % volatile (organic) matter by measuring % weight loss of the 90°C dried sample when combusted for 5 hours at 500°C. Weight loss was 5.8% and 6% for the samples.

Slime

Slime (periphyton) was scraped from the surface of a brick submerged 10–15 cm below the water surface. This slime was a mixture of algal growth (blue green Ocillatoria dominant), micro-organisms (many protozoans) and some mud. The sample was observed under 100X and 400X microscope and dried at 90°C.

Pig manure

Samples of pig manure were collected from a storage pit adjacent to pig houses on the bank of a nearby pond and dried at 90°C. Pig feed is reported to be: (per pig per day) for fattening pigs which were present until 29 June 1983:

6 jin*fresh weight of greens
1.5 jin barley
0.5 jin wheat

for piglets purchased 24 July 1983:

2 jin fresh weight of greens
1 jin wheat bran
0.5 jin waste wheat flour

The greens are mostly Alternanthera philoxeroides (the water peanut) and some Amaranth.

Vegetation

Samples of mixed species of grasses being added to the pond were taken. Only one species was identified: Eleusine indica (yard grass, a known C₄ species). The grass was dried and ground in 2 batches.

Lake Taihu had been flooded for several weeks and aquatic plants were not available on 4 August; 3 species used in the fish ponds were obtained. Vallisneria spiralis (most common), Hydrilla verticillata, and Potamogeton motainus. Samples were dried at 90°C.

DELTA C ANALYSES

Delta C analyses were performed at the Petroleum Institute of the Ministry of Geology, Wuxi.

The individual dried samples were ground in mortar and pestle. Seemingly representative portions of each sample were combusted at 850°C in an excess of oxygen in a 3-stage oven. The gas stream containing oxygen and the gaseous combustion products passed over copper oxide and silver wire (to trap sulfides), through 2–50°C water traps and then through a -196°C trap (chilled in liquid nitrogen) to freeze out the CO₂ while allowing oxygen to escape. After 4 minutes of combustion, the -196°C trap was isolated and evacuated to 10^-4 torr pressure. A sample collector also at 10^-4 torr was attached to this CO₂ purification line. The -196°C trap was heated to -50°C vaporizing the CO₂ while retaining any water vapor. Concurrently, the sample collector was chilled to -196°C. The CO₂ passed to, and condensed in, the -196°C sample collector. After 6 minutes collection was complete. The sample collector, evacutated to 10^-4 torr, was closed and allowed to warm to room temperature. The purified CO₂, formed from the combustion of the sample, contained in the collector, was analysed for the ratio ¹³CO₂: ¹²CO₂ on a Finnigan MAT 250 isotope seperation mass spectrometer and delta C calculated.

* 1 jin = 500 gm

RESULTS AND DISCUSSION OF DELTA C ANALYSES

Since carbon makes up approximately half the dry weight of most organic matter, the delta C analyses deal with the main constituents of life. Since total carbon of an organism is analysed, it is essential that this carbon be related to the system of interest. The delta C of a fish that gains only a few % of its total weight during a short time in a certain pond will not provide insight into the dynamics of that pond. Its carbon was largely accumulated prior to entering the pond. In the 2 ponds being studied all the fish more than doubled their weight during the 100 days of the experiment (Table 2). In 1982, they were grown in ponds of a similar type. In addition, during the 100 days (the elapsed time of this trial) some original carbon exchanged with the environment even in the absence of growth in these ponds.

Several aspects of the delta C data presented in Table 3 are of interest. The pig manure (delta C -28.2) reflects the fact that 3/4 of the pig diet is barley and wheat, O₃ plants (Table 1). The data for wheat, try, barley, oats of Table 1 were analyses of grains grown in temperate, often low moisture areas. The climatic difference between that and the Wuxi area might account for a shift in delta C of a few per mil. The delta C of the water peanut, the remaining 1/4 of the pig diet, has not been analysed.

The delta C of the mixed grasses (-14.1) is typical of tropical grasses which are often C₄ type. The slime on the rock taken from the one mu pond had a delta C previously reported for Oscillatoria, -14 to -15. Microscopic examination of the slime showed the blue green algae (Oscillatoria) to be the dominant species.

The delta C of the sediment after acidification to remove carbonate had a delta C (-22.3) that is typical of fish pond sediments which often fall between -22 and -24. The cause of this similarity in delta C values is not clear and may be an artifact of methods for sample preparation.

The power of the delta C analysis as an indicator of sources of fish growth is apparent in a comparison of the delta C of the filter feeders. Most obvious is (for silver carp and bighead carp) the extreme similarity of the delta C of the filter-feeders, both between species and between ponds. In the 3 cases, all delta C values fell between -28.1 and -28.9. No other groups of fish in these ponds are so similar. Filter-feeders are probably the most passive of fish in their feed selection. Size discrimination, based on gill rakers and mucus secretion, may be the only factors distinguishing between the foods used by the filtering species. In the one mu pond there was no significant quantity of seston larger than 37 microns. This is a typical observation for manure-loaded ponds having 1 or more fish per m² of pond area. The only current input to the pond having a delta C matching these filtering species is the pig manure, and the delta C match is perfect.

Numerous observations in polyculture ponds stocked with about 1 fish/m² and receiving dry chicken manure show the delta C of silver carp to match the delta C of microalgae and not the delta C of the dried manure. However, in parallel, similarly stocked ponds, when the manuring is with fluid, fresh manure such as goose manure, the silver carp delta C shifts dramatically toward the manure delta C. The flocculant nature of fresh manure appears to permit its suspension in the water column in a manner able to be filtered by the silver carp. Whether the manure provides direct nutrition or is a base for microbial growth which in turn provides the sources of growth has yet to be established.

The delta C data of these ponds indicates that the fresh pig manure has an effect similar to that of the fresh goose manure. Its fluid nature makes it available, either directly or after microbial recycling, to the filter feeding species, bighead and silver carpe.

The two grass feeding fish species (grass carp and Wuchang fish) are interesting in themselves. In both the 1 mu and 5 mu ponds, Wuchang fish delta C is significantly more negative than the grass carp delta C. The intestines of both fish are packed with chewed grass. However, delta C indicates that each species is getting part of its nutrition from a source not available to, or less exploited by, the other grass eater species. Based on the very few pieces of delta C data that we have, the Wuchang fish may be getting more of its growth from foods originating with a food web based on a C₃ source such as the pig manure than is the grass carp.

The delta C of grass carp, less negative than typical C₃ delta C values, indicates food sources based on both C₄ and C₃ species, with C₃ dominant.

The green manure added to the pond are both C₃ and C₄ types. The delta C of the grass eating species show the influence of both. It is possible (although it does not seems likely) that in their consumption of the green manures, which are thrown into the pond as bunches of whole leaves (not chopped), these fish sufficiently pulverize the leaves to make them available to the filter-feeders. This would reduce the dependence of the filter-feeders on the manure as their C₃ source. Resolution of this question is best attained by a parallel, controlled experiment with 2 sets of organic inputs to the ponds. As stated previously, such an experiment was performed using either dry chicken manure (delta C - 17) or fresh fluid goods manure (delta C -13) in polyculture ponds stocked with silver carp, tilapia (nilotica X aurea), common carp. The silver carp assimilated carbon originating in the fresh manure, while in the dry manure ponds no influence of the manure on the silver carp delta C was observed. In these ponds, however, grass was not added.

The crucian carp is considered to be an omnivorous, particulate feeder. Its delta C strongly indicates use of C₃ type foods in this pond. Its delta C is more negative than the fish species known to eat grass. This implies a food input associated with the pig manure, the only source of more negative delta C, with some growth coming from less negative sources such as the sediments (delta C -22.3).

It is interesting that the influence of the slime on the solid surfaces of the bank is not more apparent in the fish delta C. In other polyculture experiments, when Oscillatoria (delta C approx. -15) was the dominent species in slime, the slime was not observed (as per delta C analyses) to contribute to fish growth. When slime consisted of green algae and associated microbial communities (delta C typically -27), it contributed significantly to tilapia (nilotica X aurea) growth.

CONCLUSION

From this brief delta C survey of 2 ponds, we can see directions that future studies can take to reveal interactions among fish species and available foods. For the conditions of these ponds delta C has shown that silver carp and bighead take nutrition from the same source: that grass carp and Wuchang fish, though both considered to be grass eaters, take part of their nutrition from different sources; that the crucian carp is strongly influenced by the pig manure, with only slight input from other organics in the sediments or on solid surfaces.

To gain deeper understanding of the dynamics between fish and food, experiments are required having inputs with clearly defined and distinguishable delta C values. For example, to measure the impact of green manure in the presence of focal manure, the feed of the animals must be of a C-type different than the green manure. A solution would be pig feed as at present (delta C -28) with only local grasses (average delta C -14) as the green manure. This would cause minimum disruption to the man raising the pigs, while the fish ponds (1 mu, duplicate or triplicate), being under the management of the Lead Centre, would purchase local wild grass (C₄) and not rye grass (C₃) or aquatic plants. The ponds would receive, as is standard practice, both grass and pig manure.

The current experiment testing feed pellets vs snails for black carp is well suited for delta C since the pond inputs are defined. Delta C sampling of these, plus the natural foods and the black carp flesh might reveal the effectiveness of various components of the feed.

After a more complete delta C record of algal growth is established for the Wuxi Lead Centre ponds, a green manure with delta C different from the dominant algae of the area can be selected and this given to the ponds, fresh or fermented, with supplemental chemical fertilization in both cases. Delta C of the fish grown therein should clearly reveal the source of their nutrition (algae or grass). A parallel, concurrent set of ponds receiving only pig or cow manure (depending upon the delta C of each) might show the relative impact of grass vs manure on the various fish species.

Table 1 Some Approximate Delta C Values of C₃ and C₄ Plants

* Common in Wuxi and Guangchou areas.

Table 2 Data Related to Fish Stocking and Sampling at 1 & 5 MU Ponds

NOTE:	Stocking date	15 April 1983
	Sampling date	26 July 1983
	One mu = 667 m2 =
	One jin = 500 g = 0.5 kg

Table 3 Delta C values measured in 1 & 5 mu Ponds Receiving Pig Manure and Mixed Green Manures

* See Table 2 for fish species abbreviations.

** Actual analyses were on fish flesh for all species except B H. Extrapolation to whole body delta C based on measurement showing flesh delta C -1.5 = whole body delta C; and fin delta C -2.5 = whole body delta C.

Natural Foods and Added Manures in One Mu Fish Pond

NOTES: Seston almost entirely smaller than 37 microns and primarily algae.

Slime on rock had Oscillatoria as the dominant identified species.

Acidification of the sediment volatilizes carbonate so that upon combustion, the CO₂ formed originates with sediment organic matter and not sediment carbonates.

Mixed grasses (C₄) used in mid July; aquatic plants used in June - October; rye grass (C₃) used in April and May.