The design presented takes into consideration the existing situation and future needs of the pond complex for the most flexible management possibilities (Figs. 7 and 8).
There is no organized fry gathering effort, hence, fry is not available to be purchased as needed. As such, the pond complex has to be designed to enable the project to catch its own fry during the fry season and juveniles after the fry season.
The main dike facing the River Makis is low and is provided with five overflow facilities to allow water in and out of the ponds (Fig. 9). It is common knowledge that the Chanos fry voluntarily enter estuaries and ponds with the influx of the tide. The overflow devices on Dike A (Fig. 9) are provided with wooden and other suitable materials to prevent erosion due to water inflow. There is a water jump at the end of the protruding wooden channel to prevent pondfish from swimming out against the water current (Fig. 9).
The juveniles will also run up against a trickling stream of slightly less saline water than the sea or river. Thus, the pond complex is also prepared to accomplish with its long irrigation canals (AA' and BB', Fig. 11) freshwater sources as well as ponds that can act as a reservoir for continuous flow when water in River Makis and the bay is less than irrigation and pond waters height (Fig. 8).
The entrapped fry and juveniles can then be channelled into the fry and transition ponds after collecting and cleaning at the catching ponds. Ponds 1-A and 1-B are primarily fry catching ponds (Fig. 8). After the fry season, the overflow dike (A) facing the River Makis can then be sealed with boards and made water-tight with some soil, and ponds 1-A and 1-B are put to use as fry and/or transition ponds. Both can also be used as “food” and culture ponds.
With the fry catch being left more to chance, the areas dedicated to fry and fingerling (juvenile) ponds are much larger than in countries where the specific quantities of fry required are readily available by purchase.
Into the design is built a “food” Pond 3 (Fig. 8). This is a new concept tested and proven to be effective, lately. The “food” pond is an area where intensive fertilization and inoculation with micro-organisms is done and where all the other parameters (salinity, pH, etc.) are carefully controlled for optimum growth of blue green algae (fish food). The produced food is then channelled to the other ponds where the fish can utilize them, by gravity method through PVC pipes pre-laid across the dividing dikes. The food pond ordinarily does not have fish in it, but is utilized for continuous food production. However, it may be treated as just another rearing pond, if so desired. This “food” pond can also double as a transition or holding pond or part of the progression series of ponds.
It will be noted that the four fry ponds are of the same size and that the succeeding transition ponds (2E) is twice the size of the fry ponds (Fig. 8). Pond series 2 can be used for applied research on fertilization, stocking density, polyculture, etc. Replications can be made in these ponds of the same size while some of the ponds can be used as controls.
There are three main rearing ponds (Fig. 8). Pond 2-F (1 ha); Pond 4 (2 ha); and Pond 5 (4 ha). This arrangement of areas in a ratio of 1:2:4 allows the operator to use the progression method of culture (Jamandre and Rabanal, 1975). An explanation of this method is summarized briefly (ANNEX C). Using this method, the same size of stock is transferred from the transition pond into the first rearing or grow-out pond. These fish are then transferred progressively into the series of grow-out ponds doubling in area until they are finally harvested at the last pond. The transfer and/or harvest is usually done every 30 to 45 days, at which time, the food in the stocked ponds should be about depleted. The vacated ponds are again prepared for food production and would subsequently receive the new fish stock from the smaller or transition pond. The system balances quite effectively the total biomass and available food in the pond. The 30- to 45-day cycle of total drainage also affords a good opportunity for the ponds to be aerated and dried, thereby renewing its biological cycle. The system is easy to use for Chanos and mullet (Mugil), but is a bit combersome for Tilapia and shrimp due to transfer problems for the latter species. A production of 2 to 2.5 tons/ha/year is possible with the progression method for Chanos chanos, without feeding.
While the design lends itself to the progression method, it is not limited to that method alone. It can also work well with the continuous production - multi-size/mono-species method. This is where the individual rearing ponds are stocked (after preparation) with 3 or 4 size groups of fish. Usually, the beginning stock or total biomass is 50 percent of the pond carrying capacity and a new (smallest) group is introduced to make up the 50 percent of the pond carrying capacity.
Selective harvesting - harvesting under the multi-size stocking method may be accomplished in two ways:
With gillnets - the correct mesh size of the gillnet will allow the two smaller groups to go through the net and the big ones harvested.
With selective bamboo grid - Chanos chanos is like salmon because it instinctively swims against the current which may be induced by letting in water from the river at high tide. At harvest time, all the fish swim against the current into the irrigation supply canal. Then the appropriate selective bamboo screen is dropped at the necessary gate and the water flow is stopped by closing the primary gate. All the fish will try to re-enter the rearing pond, but only the small size groups will be able to enter through the selective screen. The harvestable Chanos chanos are, therefore, left in the irrigation canal which also serves as the catching basin. These are simply seined for conveyance to the market.
Harvesting every 30 to 45 days allow the food to grow back and the food is grazed off as the pond carrying capacity is approached. In this way, there is a continuous balance between stock and available food. At times, it may be necessary to supplement feed or harvest more fish prematurely in order to keep the food and stock balanced. At other times, additional small fish may be stocked when food production exceeds food demand of the biomass of the fish under culture. This may be unavoidable due to climate and other variable conditions.
The multi-size continuous culture can be managed with both shallow water (35 cm maximum) and deep water plankton method (60 cm - 1 m +) of culture. This writer has successfully practiced the use of shallow water during the dry season (to maximize use of sunlight penetrating into the pond bottom which enhances the growth of blue-green algae), and gradually shifting to the deep water method during rainy season. In the deep water method, continuous daily fertilization (5–10 kg of dry chicken manure with ½ kg 16-20-0 per ha) and micro-organism inoculation has been found to be very effective. In some cases, small-sized Chanos chanos reach growth rates of 4 g and more per fish per day. For new ponds, the following procedure may be used:
At the start of the fry season, fry are carefully cleaned of predators and stocked directly into the pond where predators have been totally eradicated with the use of bio-degrading poisons and the water has been thoroughly screened. Thereafter, the stocking rate shall be 1 500 pieces per ha, assuming a 70 percent survival in the rearing pond. After 40 days, another 1 500 fry or stunted fry from the nursery ponds will be stocked in the pond. The third stock will be introduced 80 days after the first stocking, allowing three staggered sizes at 40-day intervals. The third stock will still be 1 500 pieces with average growth of 2 g per day, the stocked fry (1 500 pieces) at day one shall have grown to 250 g by day 120. At day 40, fry should have grown to 170 g each and by day 80, the stock should have grown to 90 g each. Moreover, a 70 percent survival was assumed so the estimated first year maximum of about 500 kg pond carrying capacity will have been reached by day 120.
Text Table 1. Pond stocking and survival in multi-size culture method
| Day 1 | 1 500 stocked | 1 000 survival at 250 g | = | 250 kg |
| Day 40 | 1 500 stocked | 1 000 survival at 179 g | = | 170 kg |
| Day 80 | 1 500 stocked | 1 000 survival at 90 g | = | 90 kg |
| Day 120 | Total Biomass | 520 kg |
Since 250 g fish will be harvested at this time, the total biomass per ha after harvest is 510 less 250 kilos per ha. After harvest, another 1 500 pieces of 30 g fingerlings will be stocked in the rearing pond from the transition pond. Using the assumed 70 percent survival, the total biomass would be 305 kilos.
These routine harvests every 40 days shall go on until dry months immediately before the fry season starts. The two other sizes (90 g and 170 g) are transferred to a holding pond or canal, and the rearing pond would be totally drained, dried, poisoned, and fertilized. After food is re-grown, the 170 and 90 g fish held over stock will be re-introduced with another batch of 30 g fish.
At any time-during the culture, when the rate of growth of the fish is observed to be below average due to lack of food, it may be necessary to:
Supplement feed until such time that the natural food shall have recovered from over-grazing;
Move stock to a holding canal or pond and prepare and fertilize the whole rearing pond.
For the first year, it is estimated that five harvests will only be realized due to delays in obtaining the different sizes of fry required. Although it is possible to get nine harvests per year1, in actual practice good managers usually get a maximum of eight harvests due to the required pond preparation (drying, fertilizing, etc.) during dry season months and occasional re-fertilization of the rearing pond.
For the particular project where the people have no practical experience in the “art” of fish farming, it is safe to assume the following:
Text Table 2. Pond carrying capacity and expected yields
| Year | Maximum pond carrying capacity | No./size group survival at harvest | Kg at harvest | No. of crops | Yield/ha/year |
| 1 | 510 | 1 000 | 250 | 5 | 1 250 |
| 2 | 612 | 1 200 | 300 | 6 | 1 900 |
| 3 | 714 | 1 400 | 350 | 7 | 2 450 |
| 4 | 765 | 1 500 | 375 | 8 | 3 000 |
Note: 40 days per crop. See text Table 1.
The maximum pond carrying capacity can be checked when the stock is held in the canal. The number of each size group can be counted and an average sample weight taken. The total weight of the harvest and the stock held in the canal is the maximum pond carrying capacity.
At the Diego Suarez project, when cheap electricity becomes available at the pond site, production can easily be over the 3 metric tons/ha/year with pumps, aerators and partial feeding. The potential of 5 tons/ha/year with intensive feeding procedures can also be possible without much trouble.
Because of the presence of shrimp (predominantly Penaeus indicus, P. merguiensis, Metapenaeus monoceros and P. monodon) in the whole of Madagascar, Pond 3 is designed to have deeper water which the shrimp require. In this 2-ha area, experiments with monoculture of shrimp can be done. Shrimp/fish polyculture can also be done in this deep pond.
The canals AA' and BB' have provision for expansion should the project be expanded to full utilization of all the 80 ha in the area. (Fig.6 and 8). For the present, only the irrigation run-off freshwater is utilized, but the water of River Main may later be utilized when the project is expanded.
Presently, only 11 ha are recommended for development, but canals AA' and BB' are capable of supplying 40 ha of ponds. The planning of the 11 ha, however, takes into consideration the full utilization of the whole area of 80 ha, if later desired. Provision for expansion is already built into the 11 ha project.
It shall be noted that the design calls for one main gate for the River Makis (Fig. 13). This is a two-channel gate of 1 m per channel. Plans for a two-channel gate with provision for a propeller vertical pump are included and recommended to be adopted inspite of additional construction costs, as it will prove very beneficial later (Fig. 14).
At the centre of the 80 ha is a cross-gate of two single channels and two double channels (Fig. 15). This gate will take care of the water from the River Main (two channels in case of flooding), one channel for irrigation surface water and a double channel to the bay - to take up volume from River Main during flooding and also as a primary supply gate from the bay to the pond complex. The single channel (1 m) to the canal AA' is considered adequate.
All peripheral gates are designed for permanent construction for all of the 80 ha (Figs. 13 and 14). The present canals AA' and BB' are designed to handle up to 50 percent (40 ha) of the area, but also have been provided with expansion space. The design for secondary and tertiary concrete or wooden gates are also detailed (Figs. 16, 17, 18, 19 and 20).
As for the dikes (Figs. 8, 10, 11 and 12), it will be noted that the direction of the prevailing strong wind is from the east. The wind velocities being quite high and constant (Tables 2 and 3), the ponds are planned to be laid narrow and long running north to south so that the long north to south dikes would serve as wind breaks. Also, the narrow pond bottom (east to west) prevents the winds from developing large waves to hit the western dike edge of the ponds.
The dike berm on the eastern side of each dike is made larger than that on the western side. This is to prepare for the gradual erosion and to protect the dike against continuous wave action coming from the east.
The three dikes (GH1, GH2, and GH3) in the middle of Pond 5 (4 ha) are made simply to absorb the excess soil and also to act as wind breakers (Fig. 8 and Tables 4 and 5).
Use of high volume pumps with low heads have first been considered against large excavations (Tables 4 and 5). Because of lack of trained personnel to really intelligently maintain machineries, it was deemed wise to go into full gravity by tide operations. It is most reliable although a bit expensive at the beginning, but cheaper in the long run when compared to ponds dependent mostly on pumps.
However, the primary gate (Gate 1) is suggested to take into account the alternate design presented that has provisions for low head, high volume propeller pump to be installed when electricity becomes available in the area.
Sites for the windmill/pump and windmill/electric set up are pinpointed in the plan. It has been observed that continuous steady, strong winds are available in the area and it is strongly suggested that windmill/pump and windmill/electricity be made part of the project. A continuous windmill/pump will greatly enhance the project. It can easily make up for the high evaporation rate (20 mm/ha/day) in the area (Tables 2 and 3).
It might be mentioned that a rough estimate made by the writer seems to point to at least 1 HP net (after frictional losses, windmill efficiency and other losses) are taken into consideration. The windmill propeller sweep area for these calculations being 5 m in diameter and having a centre height of 10 m. It is suggested that if and when a windmill/pump is instituted for the project, a specially designed low head, high volume pump be coupled to the windmill, taking into account the total dynamic head not exceeding 2 m. For maximum efficiency, it is suggested that the suction for said pump be on a floating device so as to assure that the pump is operating at the lowest possible total dynamic head (TDH) under any tide situation, hence, it would have maximum water delivery at all times.
In the plan, provision is made for:
