5. FRY NURSING IN EARTHEN PONDS

5.1. Pond preparation, fertilization and feeding rates.
5.2. Impact of tadpoles
5.3. Nursing of catfish larvae in protected ponds

Four factors are of great importance when nursing first feeding larvae within earthen ponds:

· The availability of large quantities of zooplankton
· The stocking density of the 3-day-old larvae
· The duration of the rearing period
· Pond size, and ratio of dike length pond surface

5.1. Pond preparation, fertilization and feeding rates.

5.1.1. Cleaning
5.1.2. Liming
5.1.3. Fertilization
5.1.4. Daily supplementary feeding

5.1.1. Cleaning

Before the nursery pond is filled with water the banks of the dikes should be cleaned and monitored on weak points for leaks and repaired. Grasses should be cut and excess silt from the pond bottom removed. The pond bottom should then be allowed to dry for a few days so as to kill potential fry predators (i.e. water insects, amphibian larvae and catfish fingerlings from previous rearing), and to increase the mineralisation (oxidation) of nutrients in the pond bottom.

5.1.2. Liming

Liming is an important part of nursery pond maintenance, increasing the natural productivity of the ponds and having a favourable effect on the health of the fry. Some of the beneficial effects of liming can be summarized as follows:

· Disinfection of the pond bottom (using quick lime).

· Increases the pH of the water and pond bottom to an optimum level (pH 7-9) for plankton and fish production.

· Increases the alkalinity of the water; adequate alkalinity required so as to ensure pH stability and neutralize the harmful effects of magnesium, sodium and potassium salts.

· Increases pond productivity through increased biological activity and availability of minerals in the pond bottom and water column.

Quick lime:	7 - 10 kg/100 m2
Caustic lime:	7 - 13 kg/100 m2
Agriculture lime:	20 - 30 kg/100 m2

5.1.3. Fertilization

The most critical factor for the successful nursing of African catfish larvae is the ready availability of zooplankton during the first week after stocking, as they feed only on live food during this period (de Graaf et al., 1995). A good zooplankton bloom can only be obtained if the ponds are well fertilized.

In general it is believed that both phosphorus (P) and nitrogen (N) are required in minimum quantities for optimum primary production in fish ponds. The favourable action of potassium (K) has not been clearly demonstrated. Since the required quantities of these minerals are not always available in ponds, it has become a necessity to add them in order to establish an optimum standing crop of zooplankton. This can be achieved by adding minerals either directly (chemical fertilizers) or indirectly (organic fertilizers).

Chemical fertilizers

In general the mineral composition of chemical fertilizers is expressed either as a percentage of equivalent N, P₂O₅ or K₂O. In practice, the main fertilizers used are: superphosphate (containing about 20% P₂O₅), triple superphosphate (containing about 45% P₂O₅), urea (containing about 45% N) and NPK 15:15:15 (15% N, 15% P₂O₅, 15% K₂O).

Organic fertilizers

The most commonly used organic fertilizers are poultry manure, duck manure, pig dung, sheep dung and cow dung. In general, the fertilizing value of manure depends upon the C:N ratio in increasing order from cow and sheep manure followed by a grouping of pig, chicken and duck manure (Schroeder, 1980).

The quantity of organic and inorganic fertilizers required varies from place to place and from pond to pond. In the Republic of the Congo, catfish nursing ponds were fertilized with dry chicken manure at a rate of 50 kg/100 m² one week prior to stocking. This resulted in a good phytoplankton bloom, the pond water containing about 1.5-2 ml of plankton per 100 litre of water and having a secchi disk reading of 20-25 cm (de Graaf, unpublished data).

By contrast, in the Central African Republic (Janssen, 1985c) catfish nursing ponds are fertilized with:

10-20 kg manure/100 m²
0.4-0.8 kg N/100 m²
0.1-0.2 kg P₂O₅/100 m²

5.1.4. Daily supplementary feeding

Daily supplementary feeding of the catfish must start immediately after stocking (100 larvae/m²) and the feeding rate is usually 1-2 kg of rice/wheat bran per 100 m². However, before being used, the bran must be sieved through a 0.5 mm mesh. In addition to bran, the following feeding rates (seven days a week in two equal portions) are recommended for the use of formulated feeds containing fish meal as animal protein source:

First week:	0.50 kg/100 m2/day
Second week:	0.75 kg/100 m2/day
Third week:	1.00 kg/100 m2/day
Fourth week:	1.25 kg/100 m2/day
Fifth week:	1.50 kg/100 m2/day

In some cases not all the feed will be consumed completely by the hatchlings, and the remaining feed will serve as a fertilizer input and will help to maintain the plankton bloom within the pond.

5.2. Impact of tadpoles

A major problem frequently encountered is the presence of tadpoles. Tadpoles belonging to the species Rana occipitalis (Gunther 1858), Ptychadena pumilio (Boulenger 1920) and Xenopus tropicalis (Gray 1864) seem to be phytophagous and do not feed on the hatchlings stocked in ponds (de Graaf et al., 1995). By contrast, Janssen (unpublished data) found that if tadpoles of Xenopus spp. were placed together with Clarias larvae in a petri disk, the Clarias larvae disappeared within one hour.

However, the presence of tadpoles is a nuisance because they compete for the same food resources within the pond. They feed on the phytoplankton which is needed for the development of zooplankton, which in turn is needed for the growth and development of the catfish larvae.

Typical results for the nursing of C. gariepinus in unprotected ponds are shown in Table 3 and average production figures of unprotected ponds in four different locations in Africa are presented in Table 4.

Table 3. Results of nursing of C. gariepinus within unprotected ponds in Congo Brazzaville and Kenya

CONGO BRAZZAVILLE (de Graaf et al., 1995)					KENYA (Obuya et al., 1995)
No. of hatchlings stocked (No./m2)	No. of fingerlings harvested (No./m2)	Survival rate (%)	Rearing period (days)	Weight of fingerlings (g)	No. of hatchlings stocked (No./m2)7	No. of fingerlings harvested (No./m2)	Rearing period (days)
29	8.4	28.7	36	2.8	15-45	0.6	29
30	1.9	6.3	38	4.1	15-45	10.1	18
32	1.2	3.6	34	12.8	15-45	4.1	29
34	0.0	0.0	37	-.-	15-45	21.6	33
53	0.0	0.0	45	-.-	15-45	11.7	24
68	0.0	0.0	45	-.-	15-45	5.1	23
68	1.3	1.9	37	5.5	15-45	11.1	28
71	0.6	0.9	37	8.2	15-45	5	35
71	0.7	1.0	45	22.4	15-45	6.4	36
75	2.1	2.8	39	15.5	15-45	0.7	36
87	0.9	1.1	37	2.9	15-45	0.8	37
100	27.1	27.2	45	2.9	15-45	6.3	18
100	26.5	26.5	45	1.4	15-45	5.6	42
100	0	0	45	-.-	15-45	1.9	44
100	1.2	1.2	44	16.9	15-45	2.6	40
100	0	0.0	37	-.-	15-45	6.9	38
					15-45	0.7	45
					15-45	4.3	21

⁷ The exact number are not known.

Country	No of fingerlings harvested (No./m2 ± std)8	Source
The Republic of the Congo	5.0 ± 13.9	de Graaf et al., 1995
Kenya	5.8 ± 4.9	Obuya et al., 1995
Cameroon	2.7 ± 1.6	Hogendoorn, 1979
Ivory Coast	6.8 ± 7.4	de Graaf, 1989

⁸ Standard deviation

· Protect the pond against frogs and tadpoles. Surrounding the ponds with aluminium roof plates (80 cm high) proved to be very successful in the Republic of the Congo. In some countries the nursery ponds are surrounded by nylon netting (see Figure 17) but this had the disadvantage that the nets deteriorate quickly, due to the harmful effect of UV radiation/sunlight.

· Remove the bulk of the tadpoles from the pond with a small mesh-sized seine net on the day the catfish larvae are stocked, as is routinely carried out in the Central African Republic.

Figure 17. Nursery ponds for C. gariepinus surrounded with nylon nets

5.3. Nursing of catfish larvae in protected ponds

5.3.1. Stocking density of the catfish larvae
5.3.2. Size and form of the nursing pond
5.3.3. Duration of the rearing period and cannibalism among the catfish fingerlings
5.3.4. Pond monitoring and predator control

A good plankton bloom is a crucial factor, therefore the protection of the ponds against tadpoles is essential for the successful nursing of C. gariepinus in ponds.

In the Republic of the Congo (de Graaf et al., 1995) the nursing of C. gariepinus was successfully carried out in ponds which were surrounded with aluminium roof plates (80 cm). Three days after hatching, the larvae of C. gariepinus were stocked in earthen ponds (100-150 m², 0.8 m water depth) at densities varying between 70-100 larvae/m². The ponds were filled with water and fertilized with chicken dung (50 kg/100 m²) one week before stocking. From the day of stocking the fish were fed 6 days a week with wheat bran at a rate of 1 kg/100 m²/day and thereafter this feeding rate was kept constant. The average production figures obtained from 24 nursery production cycles are presented in Table 5.

Table 5. Average production figures (± s.e.m) for the nursery rearing of C. gariepinus within protected ponds from 24 production cycles in the Republic of the Congo (de Graaf et al., 1995).

Production parameter (means)	Results
Initial stocking density (No./m2)	80±5.8
Rearing period (days)	48.3±4.6
Harvested fingerlings (No./m2)	32.3±3.3
Survival rate (% of stocked total)	38.7±3.7
Weight at harvest (g)	3.1±0.5

Reliable fingerling production can be obtained from protected ponds as plankton development is not hampered as in unprotected ponds. However, two factors are of importance for the successful nursing of C. gariepinus namely: stocking density and the length of the rearing period.

5.3.1. Stocking density of the catfish larvae

For years it has been believed (also by the authors) that the optimal stocking density for larval catfish was 100/m²; harvesting about 35-40 fingerlings/m² after 5 weeks, each fingerling weighing 2-3 g each (de Graaf et al., 1995). Increasing the initial stocking density did not increase the production and lower stocking densities resulted in less fingerlings but larger fingerlings being harvested.

However, recent developments in Kenya (Campbell et al., 1995) have changed this picture, with higher stocking densities and more fingerlings per square meter being harvested as shown in Table 6.

Note:

In order to obtain the correct stocking density, the number of 3-day-old larvae harvested from the happas has to be estimated. The easiest way to do this is to estimate the number in a glass and an experienced operator can easily handle 10 000 larvae in 20 minutes. It is also possible to estimate the number of larvae volumetrically, as 1 ml contains about 150 larvae. The disadvantage of this method is that the larvae can be damaged which results in high mortalities after they are stocked in the nursery ponds.

Table 6. Results of nursing C. gariepinus within earthen ponds (10 × 10 × 1 m, water depth 30 cm) completely covered with a nylon net (mesh size 4 mm, Campbell et al., 1995).

Parameter (mean)	Results ± s.e.m
Initial stocking density (No./m2)	278±17.7
Rearing period (days)	18.4±1.2
Harvested fingerlings (No./m2)	85±11.6
Survival rate (%)	32.2±3.9
Weight at harvest (g)	0.3±0.07

Stocking densities as high as 250 larvae/m² with an average production of 85 fingerlings/m² have only been obtained before in Africa by Janssen in Nigeria (unpublished data). In South Africa nursery ponds are repeatedly stocked at a rate of 2 000 fry/m² and about 500-800 fingerlings/m² are harvested (Hecht et al., 1988). However, these ponds are stocked with 10-day-old fry (20-30 mm) and so this cannot be directly compared with the stocking of hatchlings.

Although these results are only preliminary, it is worth noting that the larvae are reared in very small (10 m²) shallow ponds which were well fertilized and completely covered with a nylon net (mesh 4 mm). The small size or the high dike-length/surface area ratio of these ponds could be of particular importance and is currently being further studied in Kenya.

5.3.2. Size and form of the nursing pond

In general, the production and survival of C. gariepinus has been found to be higher within small ponds rather than in larger ponds; smaller ponds being easier to manage and to fertilize, so as to develop a plankton bloom rapidly and so ensuring plentiful food for the stocked hatchlings.

The hatchlings in Kenya are only reared for about 18 days and to an average weight of about 0.3 g. However, the survival rate of 30-35% is comparable to that obtained from protected ponds in Brazzaville. It is interesting to note that the survival rate obtained after 14-15 days of rearing (29.8±11) did not differ greatly from the survival rate obtained after 22-28 days of rearing (36.0±24.6). This indicates that the major mortalities were occurring during the early period of the nursing phase. The critical factor is most probably the availability of plankton during the first days of the rearing cycle.

Another factor which could be of importance, and which is currently being studied by an FAO project in Kenya (Support to Small Scale Rural Aquaculture in Western Kenya, FAO/TCP/KEN/4551), is the form of the pond and in particular the ratio between the dike length and the water surface area.

During the first days of the rearing period, the hatchlings can be seen wriggling in the water layer near the embankment. The hatchlings could be attracted by the shelter provided by the grass/weeds on the embankment or could possibly be attracted to the additional food being available in this area. In Table 7 the major production parameters of protected nursery ponds with different dike-length/surface ratios are compared.

Table 7. The Production figures of protected nursery ponds with different dike/surface ratios

Production parameter (mean ± s.e.m)	Protected nursery ponds in the Republic of the Congo. Dike-length/surface area = 0.5 (de Graaf et al., 1995)	Protected nursery ponds in Kenya. Dike-length/surface area = 2.2 (Campbell et al., 1995)
Initial stocking density (No./m2)	80 ± 5.8	278 ± 17.7
Rearing period (days)	48.3 ± 4.6	18.4 ± 1.2
Harvested fingerlings (No./m2)	32.3 ± 3.3	85.0 ± 11.6
Survival rate (%)	38.7 ± 3.7	32.2 ± 3.9
Weight at harvest (g)	3.1 ± 0.5	0.3 ± 0.07
No. of fingerlings at harvest per cubic meter of pond water	40.3	283
Standing crop at harvest (kg/ha)	1 250	850

It could be argued that the more embankment available, the better the results of the nursing phase, as the production is more than doubled in ponds with a high dike/surface ratio. However, ponds with a lower dike/area ratio supported a higher standing crop at harvest, which is logical as supplementary feeding is the major food source toward the end of the nursing phase. It follows, therefore, that the major benefit of a high dike/surface ratio would be during the first part of the nursing phase, when the hatchlings mainly depend upon plankton as their main food source.

5.3.3. Duration of the rearing period and cannibalism among the catfish fingerlings

A major factor which influences the success of the nursing phase is the length of the rearing period. In general, four to five weeks after stocking, two distinct size groups of catfish can be recognized within the pond (see Figure 18):

· A large group (80-90% of total biomass) consisting of small-sized fingerlings (2-3 g).

· A small group of fingerlings (10-20% of total biomass) consisting of large-sized fingerlings (8-10 g).

Note:

If 100 larvae/m2 are stocked it is essential to harvest your fingerlings 35-40 days after stocking otherwise the larger fingerlings will predate on the smaller ones.

Figure 18. Frequency distribution of harvested C. gariepinus fingerlings nursed in an earthen pond (de Graaf and Campbell, unpublished data) or nursed in a 2 000 litre circular plastic pool (Hecht et al., 1988).

Nursing in ponds, Kisumu, Kenya de Graff & Campbell, non published data

Nursing in a circular plastic pool

Source: Hecht et al., 1988

5.3.4. Pond monitoring and predator control

Within unprotected ponds tadpoles must be seined out and removed prior to stocking. Aquatic insects such as water-scorpions, water-beetles, water-boatman and dragon fly larvae can be controlled with kerosene (0.5 l/100 m²). The kerosene should be poured along the windward side in the early morning of the first rearing day. The breeze will carry the fuel across the water which will prevent the aquatic insects from taking air at the surface and they will soon die.