Mr. F. GHION
It is only in the last few years that the rearing of mullet is being carried out while employing reliable methods. This practice was derived from the standards required in “valliculture”. With the intensive rearing of fish the first year, a means to increase the steady rate was remarked with this practice. In addition to this, the integrated valliculture technique was employed, in which, mullet play a principle role. In turn, and following the availability of food suitable for this species, the rearing of mullet in fresh water became developed, while limited to the 2 species, which in such an environment show a more rapid growth rate: Mugil cephalus and Liza ramada.
This slow development was backed by a considerable series of experimental tests, carried out for the formulation of a diet, especially suited to mullet, which up to this had not existed.
This text gives the history of this evolution along with the synthesis of the research activity which has been carried out over the past ten years.
MULLET IN TRADITIONAL VENITIAN VALLICULTURE
I assume that you already have a knowledge of the rearing practice involved by valliculture techniques which programme more or less similar interventions for all the species reared and so those applicable for mullet.
Thus there exists for mullet, seeding in the Spring (the fry is either bought or obtained from the stock), growth in summer in the “Valli” tanks, direction back and capture in the estuary which is followed by the distribution into the wintering tanks. This cycle is repeated each year until commercial size is reached.
In fig.1, can be fount the 5 species of mullet deviated into our “valli”. Beside the name of the fish, the size and the commercial value are indicated, and as can be remarked these vary depending on the fish species.
Figures 2 and 3 show the growth variations in intensive valliculture. These graphs represent the general values. The growth can very from one valli to another and also from one season to another. It is however evident that 3 years of growth is necessary for a species sold at the smallest acceptable commercial size (ex. 300 g. L. aurata) and a longer time for the others.
As already remarked here above, intensive rearing of mullet was introduced in the aim of obtaining a superior steady rate later on.
Indeed, we though, that the rearing of fry in small tanks for several months would permit the distribution in the extensive tanks, of fish of greater size, which would be capable of defending themselves and escaping carnivorous fish (fig. 4).
A diet especially suited to mullet was unfortunatly not available, so the food given normally to trout and other fish was employed.
By simply taking into consideration the natural food of mullet, it is clear that this food if not incorrect was, at least, inadequate.
In the food chain, mullet is placed indeed at a quite low trophic level. In examining fig.5, which represents the components of the aquatic ecosystem, it can be said that mullet is located at primary consumer level while invading somewhat, all the same, the secondary consumer zone (carnivorous).
Litterature indeed recalls that mullet is principally detritivorous and phyto planktonphageous but occasionally, it can ingest small organisms. From an energetic view-point while admitting that for each ring of the food chain only 10 % of the energy used is restored, it seems evident that mullet, when compared with the other fish in the “valli”, makes better use of the primary energy produced by the real ecosystem. This better use can be estimated at around 10 at 100 times more than that of the other fish.
The natural diet of the mullet which also comprises a great quantity of inert material, thus appears as a poor diet. In general, it would seem correct to believe that artificial food should follow the same pattern.
The most important point in a diet from a commercial view-point is represented by the protein rate.
A series of research tests have been carried out so as to define the minimum protein rate of the food for mullet, and the summary of these is given here following :
Figure 6 shows the growth variations of fry (M. cephalus), submitted to diets of variable protein content (from 20 to 80 %). The result obtained after 20 weeks of rearing in an aquarium are shown in table 2. The comparison is easier in fig.5. The minimum content is found around 40% diets containing 20 to 30 % shows an advantage obtained at growth level while diets with high protein content show their complete futility.
Other tests have been done with C. labrosus, L. aurata and L. ramada, while using 2 different diets (see fig. 7). The essential difference between A and B diets is the carbohydrate rate content, the lipid content is nearly always constant and that of proteins varys from 10 to 50% (see fig. 8).
Type A diet which presents a Lack in carbohydrates did not permit to establish a minimum protein level (M.P.L)while for type B diet (with more sugar content) the M.P.L was reached for each species, at around 20 % (fig. 9).
Other tests of the same type, trying to establish the influence of the lipid content or the origin of proteins, have given to conclude that a diet for mullet must contain a rather low protein rate (20 to 30 %), a high carbohydrate rate (60%), and a lipid rate of between 5 to 10 %. It is clear that such values can be reached if plant life diets are employed.
Intensive rearing of mullet is thus developed by following these instructions.
The preparation of a food can not however be limited to its simple formulation. In the special case of mullet, the importance of the granulometry of the natural food is know. It is also known that mullet can adopt very diverse food techniques by adapting themselves to the characteristics of the substratum. Thus, mullet feed by sucking and filtering the bottom layer, by scraping the patina covering the submersed material, and by sucking up the tiny fragments of floating organisms.
Figure 10 shows the results obtained in a test carried out to reveal the quantity of food eliminated by the mullet, depending on the size of the particles offered. With this method it is noticed that 40 %of the food commercialized on the market is eliminated.
It must be remarked that some particles of this food are bigger than 1 000 um. There is less waste depending on the number of big particles eliminated. Furthermore, the chemical analysis carried out on this food eliminated, shows that the use of the protein part is at it's best, when the size of the particles is less than 250 um. This can be also due to the fact that the protein part of the food is generally represented by fish meals or finely minced meat. However, while analysing all the results obtained by the both tests, it seems clear that a finely minced food causes less wastes and is thus finally more interesting.
The following figure shows the results of an experiment carried out so as to verify the efficiency of the filtration systems of mullet.
For this test, 5 diets of different granulometries were prepared (from under 50 to 1 000 microns) together with 5 lots of inert material (sand). Each separate diet was mixed with one lot of inert material in every possible combination, while using around 30 % of food and 70 % of inert material. The organic substance content of the diet obtained was then defined. After the food had been given to the fish were sacrified and the analysis of the organic substances found in the stomach was established.
The variation between the organic substance content of the food and that of the stomach was chosen as indication of the ingestion capacity.
From the results of fig. 11 and 12, it appears when too fine material is presented, the mullet is incapable of selecting out the organic material, while it is quite capable of doing so when the particles are of between 100 to 250 um. Thus it can be concluded that these are the ideal dimensions for mullet food.
As for the distribution of food, a test was carried out in concrete tanks where the food was distributed in the following manner :
Whole food distributed automatically over the water surface.
Food mixed with sand distributed automatically over the water surface
The same but with adult fish.
Food mixed with water and placed on the bottom.
Food mixed with water and sand, placed on the bottom with adult fish.
Food mixed with the mud of the “valle” and placed at the bottom.
For this experiment, juveniles weighing on average 10 g. for all species were employed.
The results are found in fig.13. It can be remarked that the best growth is obtained with the first system which also gave the best indication of conversion. The use of sand seems to have little influence on growth apart from in the case where adult fish are mixed with juveniles. No advantage is obtained when the mud from the valle is mixed with the food and this confirms the observation made here above on how mullet are uncapable of selectiong out very fine material (like the clay employed).
These observations permit us to arrive at a new conclusion concerning the feeding of mullet. It seems interesting to distribute whole food over the water surface. From a practical view-point, this technique is very simple.
By applying these results, the intensive rearing of mullet can become widespead.
It is limited during the first and second year in valliculture and widespead in fresh water rearing stations until commercial size is reached, this is obtainable in two years.
To recall the observations made here above on the natural feeding habits of mullet and the ecologic value (energetic saving) that this species represents, it can be said that our final objective can not only limit itself to producing an artificial food but it seems more important to do research work which would increase the production of mullet while making better use of the natural environment.
it is evident that in this sector, the problems appear to be more complex as they concern the natural procedures which we would like to channel in a predifined direction. The idea of wanting to increase the primary production through fertilization, work, etc…is valid and applied correctly in fresh water aquaculture. This appears more difficult in a sea-water environment and especially complex when we refer it to valliculture.
Let us examine fig.14 which shows the frequence of occurance of organisms found in the stomach of 40 mullet of different species captured in a pilot facility of integrated valliculture. This figure brings nothing new to the feeding habits of mullet which are already well known. The recent presence of all the animal species already described, abundance of organic and inorganic detritus, phytoplankton and plant life.
Fig. 15 shows the situation of one species: L. ramada captured in sea-water. The selection of small material seems quite evident. The same species captured in fresh water shows a real preference for plant organisms along with rotifers which were evidently in full growth at the time of capture.
This confirms on one hand the adaptability of the species to the most varied conditions of the environment from a Trophic viewpoint and on the other hand it also shows how mullet can take an immediate advantage from the flow back of an intensive facility thus also anticipating the fertilizing effect. It can also be affirmed that in this case, the mullet play a depuration role for the water by eliminating the suspended matter.
As for the ploughing machine used in valliculture, it seems that this has already been referred to. I would only like to add, that experimentally, 1 t/ha of yearold mullet was reared in a pond of around 4 ha, by fertilizing the environmentwith 18 t of organic matter for the whole season and an intervention by means of the machine was employed every fortnight.
Apart from the high concentration of fish observed (estimated) it was verified that the intervention of the machine helped get rid of sulphuret and phosphorus contained in the mud at the bottom. This ploughing permits on one hand better hygiens in the bottom and the water and on the other hand recirculates a great quantity of fertilizer which induces a better phytoplanktonic production (increase of the chrolophyl in the water).
The establishment of the exact technique of menagement for the ploughing machine and the possibility of introducing organic matter into the environment remains a problem to be looked into. The careful use of the suggestions dictated by common sense together with the knowledge of the feeding habits of mullet introduce the development of the semi-intensive rearing of mullet which gives good hope for the future.
It is also clear that the industrial production of this species can not be performed until sure and cheap methods for wintering can be ensured.
Fig. 1. Specific characteristics of the different species of mullet (adults)
Fig. 2. Growth curves of mullet in extensive valliculture
Growth curves classed as “good” for the extensives of Venitian “valle”. In the different “valli” the rearing cycle must be prolonged so as to obtain these results. (Seeding is carried out in March – April using fry of 2 – 4 cm. These graphs do not permit the definition of the decrease in weight during winter.)
Fig. 4. Comparison between the survival percentages obtained in intensive and extensive rearing (first year) of the different species reared in valliculture.
Fig. 5. Drawing of a “pond” ecosystem. principal factors are:
I. Organic and inorganic abiotic substance
II. A. Generators - Vegetation with roots
II. B. Generators - Phytoplankton
III. 1 A. Primary consumers (Herbivora) - On the bottom
III. 1 B. Primary consumers (Herbivora) - Zooplankton
III. 2. Secondary consumers (Carnivora)
III. 3. Tertiary consumers (Secondary carnivora)
Fig. 6. Ponderal growth of M. cephalus fry reared while employing diets of different proteinic content
Fig. 7. Ponderal increase of 0+ mullet fed for 20 weeks diets of different proteinic content.
Fig. 8. Comparison between the ponderal increases of the different species of mullet (fry) reared while employing diets of different glucidic and proteinic content.
| PROT | LIP | EST.INAZ. | ||||
| A | B | A | B | A | B | |
| 10% | 9.88 | 10.66 | 10.21 | 9.87 | 6.83 | 76.07 |
| 20% | 20.18 | 22.28 | 11.06 | 9.64 | 11.87 | 64.68 |
| 30% | 30.39 | 32.67 | 10.46 | 9.19 | 17.16 | 51.25 |
| 40% | 43.00 | 44.15 | 10.24 | 10.06 | 19.85 | 38.56 |
| 50% | 54.00 | 55.27 | 10.20 | 10.78 | 26.04 | 29.70 |
Fig. 9. Composition of the different diets employed in the experiment shown in fig. 8.
Fig. 10. A graph showing the percentage of waste from the food given to the mullet, depending on the size of the particles (above) and the percentage of use in proteins, also depending on the size of the particles of food (below).
Size of the inert particles ( μ)
| < 50 | 50 – 100 | 100 – 250 | 250 – 500 | 500 – 1 000 | ||||||
| < 50 | - 14.6 | - 27.7 | - 67.9 | +114.0 | + 60.1 | - 28.7 | - 29.7 | - 42.4 | ||
| - 56.2 | - 6.6 | - 61.6 | - 2.0 | - 48.5 | - 26.6 | - 58.1 | + 90.0 | |||
| - 21.8 | - 10.1 | - 64.1 | - 54.1 | - 64.4 | - 61.8 | + 87.5 | + 16.9 | |||
| - 14.8 | - 15.3 | - 57.4 | - 63.7 | - 48.9 | + 56.5 | + 10.4 | ||||
| - 23.5 | - 23.3 | - 18.3 | - 38.2 | - 39.6 | ||||||
| - 27.3 | ||||||||||
| 50 – 100 | + 40.2 | + 26.4 | - 29.2 | + 22.7 | + 115.5 | + 7.1 | + 71.6 | + 72.6 | ||
| + 37.1 | + 34.9 | - 32.2 | + 0.4 | + 57.9 | + 100.4 | + 193.1 | + 213.9 | |||
| + 28.6 | + 36.8 | - 25.8 | + 003.9 | + 165.2 | + 43.4 | + 182.5 | + 175.1 | |||
| + 31.7 | + 31.9 | - 9.0 | + 35.7 | + 382.3 | ||||||
| + 22.7 | ||||||||||
| 100–250 | + 59.9 | + 62.8 | - 11.4 | - 6.5 | + 70.2 | + 9.8 | + 53.3 | + 72.2 | + 135.7 | + 76.3 |
| + 96.6 | + 103.7 | - 17.3 | - 22.7 | + 163.7 | + 174.9 | + 123.8 | + 150.7 | + 134.4 | + 118.9 | |
| + 95.1 | + 67.1 | - 41.4 | - 42.4 | + 32.1 | + 67.8 | + 200.5 | + 10.7 | + 97.0 | + 210.2 | |
| + 121.8 | + 0.4 | - 30.8 | - 23.5 | + 68.9 | + 111.8 | + 163.4 | + 186.4 | + 176.5 | + 117.8 | |
| + 239.4 | + 4.3 | - 9.0 | + 139.3 | |||||||
| + 43.8 | + 184.5 | |||||||||
| + 73.3 | ||||||||||
| 250–500 | + 40.9 | + 46.4 | - 56.4 | - 50.4 | + 87.1 | + 66.6 | + 38.6 | + 169.6 | + 45.7 | |
| + 54.7 | + 45.4 | - 44.5 | - 0.8 | + 188.9 | - 28.0 | + 44.9 | + 195.9 | + 123.1 | ||
| + 21.2 | - 16.0 | - 42.8 | + 55.3 | + 61.3 | + 79.6 | + 241.7 | + 253.2 | |||
| + 70.9 | - 49.3 | - 27.0 | + 20.6 | + 41.3 | + 128.5 | + 253.2 | ||||
| + 24.5 | - 77.5 | - 47.5 | ||||||||
| 500–1 000 | + 64.6 | + 46.4 | + 68.4 | - 32.6 | - 4.0 | + 155.6 | + 83.4 | + 12.3 | ||
| - 49.5 | - 16.0 | - 12.0 | - 63.3 | + 31.1 | + 83.8 | - 15.9 | + 71.7 | |||
| + 2.3 | + 45.4 | + 64.0 | - 14.6 | + 13.5 | - 11.5 | + 104.9 | + 110.0 | |||
| - 18.5 | - 49.3 | - 53.2 | - 55.0 | + 8.7 | + 126.2 | + 139.3 | + 201.1 | |||
| + 43.4 | + 63.8 | + 108.2 | + 161.5 | + 138.7 | ||||||
Size of the particles of the food
Fig. 11. Experimental data of a test carried out in the aim of verifying the efficiency of the filtration system of mullet.
Size of the particles of the food (μ)
| < 50 | 50 – 100 | 100 – 250 | 250 – 500 | 500 – 1000 | |
| < 50 | - 213 | - 62.7 | - 16.1 | - 39.0 | + 53 |
| 50 – 100 | + 33.4 | - 24.1 | + 40.7 | + 81.6 | + 184.3 |
| 100 – 250 | + 76.3 | - 24.5 | + 95.7 | + 120.1 | + 119.7 |
| 250 – 500 | + 42.4 | - 18.3 | + 75.2 | + 51.1 | + 116.6 |
| 500 – 1000 | + 3.2 | - 6.1 | + 22.6 | + 92.5 | + 100.7 |
Size of the inert particles (μ)
Fig. 12. Table of the results (average value) from the tests shown in fig.4
| Feeding system | Average weight | Average weight increase (%) | Conversion Index | ||
| Init. | Final | ||||
 | Dry food on the surface - automatic feeder | 9.8 | 24.1 | 145.9 | 2.42 |
| The same as above + sand | 10.4 | 32.7 | 214.4 | 2.73 |
| 72.9 | 180.4 | 149.2 | |||
| The same as above with small and big fish | 12.8 | 29.8 | 132.8 | 2.82 |
 | Paste on the bottom | 9.7 | 19.5 | 101.0 | 3.32 |
| Paste with sand on the bottom with small and big fish | 9.9 | 20.0 | 102.0 | 3.72 |
| 59.0 | 128.3 | 117.4 | |||
| Paste on the bottom + mud | 13.3 | 27.8 | 109.3 | 4.25 |
Fig. 13. Data from a test carried out the aim of verifying the influence of the feeding systems on the growth and conversion index.
Fig. 14. Frequences of the distribution of prey for mullet depending on their size.
Fig. 15. L.ramada: Sea water
Fig. 16. L.ramada: Fresh water
Fig.17. Sea bass intensive rearing discharges
