Previous Page Table of Contents Next Page


The International Biological Programme (1964/74) included in its work investigations into “those unused, and little used, plant and animal products which might be converted into nutritionally exploitable materials by bio-engineering techniques.” Three international meetings were held, all on the subject of novel protein resources. The last, held in India in 1970, was specifically on leaf protein. The proceedings were edited by N.W. Pirie and published in 1971 as IBP Handbook No. 20, “Leaf protein: its agronomy, preparation, quality and use”. Pirie, who is well known for his extensive work on the subject, wrote two of the chapters, surveying in detail experiments and equipment on protein extraction which include data on water plants (see analyses cited in Chapter III) and 17 of his papers, up to 1970, are listed in the bibliography. A detailed review by C.E. Boyd in the same publication is mentioned in the references given below.

For protein analyses of many aquatic plants see Chapter III.

Boyd, C.E., 1968 Fresh water plants: a potential source of protein. “Econ.Bot.,” 22:359–68

The author recorda the protein extractability data and the protein yield for a range of aquatic plants. He also mentions the ease of protein extraction. The summarized data is tabulated below:

Species% Pulp N extracted as protein NComment1
Justicia americana29.5–53.6A
Sagittaria latifolia25.2–27.4A
Alternanthera philoxeroides35.8–42.6B
Nymphaea odorata36.2–61.0B
Orontium aquaticum41.9–43.8A
Jussiaea decurrens34.2C
J. peruviana35.0C
Brasenia schreberi35.8C
Elodea densa33.6A
Nuphar advena21.8C
Polygonum sp.20.3C
Nymphoides aquaticum16.1B
Nelumbo lutea13.9A
Hydrocotyle sp.28.1A
Ceratophyllum demersum32.2A
Myriophyllum braziliense14.5A
Najas guadalupensis39.4A
Hydrodictyon reticulatum27.4A
Spirogyra sp.20.4A
Pithophora sp.15.2A

1 A. No difficulty in grinding or making extract.

B. No difficulty in grinding, contained slight mucilage butextracted well.

C. Contained excessive mucilage. Difficult to grind and extract.

Commenting on these results Boyd points out that they were ideal, laboratory, figures based on minimal losses during processing. In addition, slightly more protein was precipitated from the extract by trichloracetic acid than by heating to 80°C, which is the method of coagulation used in large-scale production. Laboratory efficiency is probably not possible with large-scale machinery, but the data are indicative of species which lend themselves to extraction.

Boyd, C.E., 1971 Leaf protein from aquatic plants. In IBP Handb., (20):44–9

The author points out that exploitation of aquatic plants for leaf protein will present unique problems in each locality, depending on species availability, harvest technique; environmental conditions and various other factors. Therefore development of this resource must be on a regional or local basis. In addition it may often be more practical to test various species by trial and error rather than follow an elaborate experimental approach. He suggests a programme which may be useful in utilizing aquatic plants as a crop for leaf protein extraction:

Development of a leaf protein resource from aquatic plants

  1. Small-scale screening of local species

    1. Extractability of protein

    2. Yield of protein per unit area

    3. Local abundance of promising species.

  2. A. Large-scale extraction trials
    B. Research on harvesting
    1. Harvesting equipment

    2. Timing of harvest for maximum protein yield

    3. Regrowth

    4. Propagation

    5. Ecological studies of promising species

  3. Nutritive value and acceptance of protein

  4. Cost analysis

  5. Bulk production

Boyd, C.E., 1974 7. Utilization of aquatic plants. In Aquatic vegetation and its use and control, edited by D.S. Mitchell. Paris, Unesco, pp. 107-14

After briefly reviewing work on protein extraction, the author records the amounts obtained from a few species suitable as raw material for protein. He found that the leaf protein from aquatic plants was similar in chemical composition to leaf protein from crop plants.

SpeciesLeaf protein yield
Justicia americana590
Alternanthera philoxeroides478
Sagittaria latifolia362
Nymphaea odorata197
Polygonum sp.180

He adds that research should be extended to other species.

Cifuentes, J. and L.O. Bagnall, 1976 Pressing characteristics of water hyacinth. J.Aquat. Plant Manage., 14:71–5

This paper is concerned primarily with the problem of pressing water out of the plant to facilitate harvesting (see Chapter IV). The authors found that juice loss during pressing increased as plant particle size decreased, as indicated below (Carver pressing cage at 1100 kPa):

TreatmentFluid expressed (%)
Whole plants17.5 ± 6.4
Chopped 25-mm pieces23.8 ± 5.1
Minced52.7 ± 3.1
Thoroughly blended57.5 ± 15.6

Thus water hyacinth should be minoed, or otherwise very finely divided, before or during pressing to maximize water removal. Such reduction destroys the fibrous structure of the plant allowing the pressure to be applied to the fluid, and damages cells, exposing the fluid for random drainage. Filtrate opacity increased as particle size decreased, indicating higher loss of solids. Juice expression increased rapidly with increasing pressure up to 600 kPa, but higher presure did not remove much more juice. Agitation of the residual cake between additional pressings increased juice. Agitation of the residual cake between additional pressings increased juice expression by 10–20%. The protein content of the juice is shown in the following table:

Analyses at pressures ranging from 172–1380 kPa

SampleDry matter (% wet)Crude protein (% dry wt.)
Liquid1.05 ± 0.2066.4 ± 12.8
Solid9.57 ± 0.9318.0 ± 0.9

Note: None of the analyses correlated significantly with pressure.

*Datta, R.K., et al., 1966 Leaf protein - preparation of protein concentrate from leaves of water hyacinth. Sci.Cult., 32(5):247–9

Of various methods of comminuting the leaves a blade-type hammer mill was found to crush satisfactorily both tender and fibrous leaves. Various extracting agents were used at the rate of about three times the weight of agent to the crushed mass. Stirring took place for 30 minutes. After filtering (by centrifuge) the protein was precipitated by bringing the pH of the filtrate to 4.0 with commercial hydrochloric acid, followed by heating to 80°C. The results are tabulated below.

Extracting solutionExtraction time (min.)% Total protein extracted (N × 6.25)
Ether-saturated water307
Sodium chloride 2%3012
Sodium chloride 5%3015
Sodium carbonate 0.5%3063
Sodium carbonate 1%1557
Sodium carbonate 1%3064
Sodium carbonate 2%1560
Sodium carbonate 2%3077
Sodium carbonate 2%6080
Sodium carbonate 4%6069

As a result of the above tests a standard extraction was selected of 2% solution of aqueous sodium carbonate, for 30 minutes.

Tests were also made of the effect of pH on efficiency of protein precipitation.

Resultant pH of the extract% of protein precipitated (N × 6.25)

Thus pH 3.8–3.5 was chosen as being the most effective range for maximum protein precipitation.

It was noted that heating to 100°C of the extract without acidification brought about coagulation of the protein to the extent of 70 to 80%. Prior acidification to pH 3.8 is not only helpful in affecting coagulation of most of the protein of the extract, but made the coagulum large enough to allow quick setting.

Centrifuging the slurry removed adhering hydrochloric acid, sodium chloride, and some disagreeable odour and taste. At this stage the product looked yellowish green.

When drying the protein in a tray, at 60°C, in a current of air, it was found that case hardening of the protein cake, which slowed the process, could be avoided if the moist material was washed with a little alcohol. When drying in the sun, or at a high temperature, the green-coloured protein changed to brown.

Koegel, R.G., et al., 1973 Increasing the efficiency of aquatic plant management through processing. Hyacinth Control J., 11:24–30

The authors describe mechanical methods of processing aquatic plants in order to reduce the quantity of water they contain for subsequent easier disposal. The plant they worked with was mainly Myriophyllum spicatum. They show that a well designed screw press may reduce the moisture of plant material by 60–65%, reducing the weight and volume of the fresh plant material to one third and one sixth of the original values. At the same time less than 20% of the nutrients in the plant material are lost in the expressed liquid.

They found that heat treatment facilitated mechanical dewatering and also caused protein denaturation or coagulation. The protein molecules then lost their water binding ability. This coupled with the inability of the cell wall to maintain osmotic pressure, or possibly even rupturing the cell walls, facilitates the passage of the liquid. Coagulation causes entrapment of the cellular matter and hence results in smaller amounts of this matter being lost with the liquid.

*Oyakawa, N., W. Orlandi and E.O.L. Valente, 1965 The use of Eichhornia crassipes in the production of yeast, feeds, and forages. Proc.Int.Grasslands Congr., 9(2): 1707–10

The authors state that water hyacinth has a promising food potential. But as it is relatively poor in protein content they studied how this deficiency could be overcome. They worked on a method of growing yeast on the macerated weed, as follows:

  1. wash the plant thoroughly, including the roots;

  2. total grinding in a liquifier or grinding mill;

  3. sterilization in an autoclave for one hour, at one atmosphere;

  4. juice extraction, while the material is still hot, through a press;

  5. centrifuge on press filter;

  6. add plain water in a quantity equivalent to the volume of the juice obtained (approximately 6 litres of water to 10 kg of fresh plant);

  7. mix both liquids to obtain a density of 1.009 at 20°C.

This is the culture medium for the multiplication of selected micro-organisms. It has a composition of:

reduced sugar and glucose0.85%
total solids3.8%

A yeast found growing on the surface of decaying water hyacinth was chosen. It was Saccharomyces cervisiae var. ellipsoideus.

Growing method. Growing on a semi-industrial scale is done in a fermenting vessel with a capacity of over 20 litres equipped with an aerator, an agitator, temperature control, flow rate control of the broth and sample collector. The material of the vessel should permit sterlization and cleaning. At the beginning the vessel is filled to 25% of its volume with sterile broth of water hyacinth, the pH of which must be corrected to 4.4 by using H2SO4. Then the media is inoculated with 20% volume with a suspension which must contain at least 50 x 106 cells/ml of the yeast. At the same time, the broth is agitated and aerated with sterilized air at not less than 5 litres/minute for each 10 litres of liquid. This aeration is necessary to avoid an undesired production of alcohol which will reduce the yeast yield. The temperature is maintained at around 33°C. After five hours the vessel is filled to capacity with sterilized broth of water hyacinth. Agitation and aeration are still necessary.

Six hours after the vessel has been filled (i.e., after a total of 11 hours), the velocity of cell multiplication will reach a maximum of a growth logarithmic curve which corresponds to a population of two billion cells/ml., equivalent to the origin of one cell from another in two hours time. During this phase a constant quantity of yeast can be obtained simply by substituting every two hours about 25% of the juice already fermented by another amount of sterile broth. Obviously, all the conditions described before should be well maintained in relation to the aeration, pH, etc. The production may continue indefinitely since contamination by foreign micro-organisms is avoided, which may easily be maintained for the first 15 days. Every phase of growing should be followed by direct microscopic checks which will indicate multiplication, contamination, etc. The juice from the propagator contains in suspension the yeast cells which may be separated by filtering or preferably by centrifuging, from which the consistency obtained is creamy, with approximately 35% of total solids. This yeast cream has a very agreeable odour and flavour, and may even be used for human consumption It may be dehydrated and ground, giving a powder of brownish colour with the same properties already described and of the following chemical composition:

 % %
mineral salts7  

If concentrates for animal feeding are wanted, the dry and ground yeast is utilized only as a source of protein and vitamins, and the dehydrated water hyacinth bagasse is utilized, resulting from the broth fermentation , as a source of carbohydrates and mineral salts. The final product has a good appearance, with good volume and its odour is similar to that of the fish meal. It may also be used in concentrates.

The yield of yeast based on total sugars in the juice varies from 160–210% for the wet yeast with 35% of solids and 42–50% for the dried yeast with 8% of moisture. On an average, each kilogram of fresh plant yields 12 g of pure dried yeast with 52.7% of protein. The yield of the bagasse with 8% of moisture is about 3.5%. As an average, 1 kg of fresh water hyacinth may produce 42 g of a mixture of yeast and bagasse with a minimum of 28% of protein.

Considering that this plant contains an average of 96% water and that 1 kg of plant material may yield, in 11 or 12 hours, 12 g of yeast with 52.7% of protein, or 6.32 g of pure protein, we get a high rate of plant conversion into high quality protein, calculated as 6.5:1 per hour which never can be obtained by the digestion of any efficient herbivore. The protein obtained can be classified according to its analytical values as between that of egg protein (100%) and leguminous protein (50%) with an average of 87%.

Taylor, K.G., R.P. Bates and R.C. Robbins, 1971 Extraction of protein from water hyacinth. Hyacinth Control J., 9(1):20–2

The authors decided it was not practicable to extract protein only from the leaves under the harvesting system used to remove the waterways, so they worked with the whole plant using two different methods.

  1. A Hiller harvesting machine chopped up the plants into coarse particles which were then passed through a screw press. The juice obtained was clarified in a centrifuge, then acidified with concentrated hydrochloric acid to pH 3.8, and heated to 80°c. It was refrigerated for 36 hours at 1°c after which the precipitate was filtered off.

  2. Small samples of whole hyacinth plants (1 kg) were chopped and then comminuted in a 4-litre capacity blender with 3 kg of 0.05 N sodium hydroxide (NaOH) at low, medium and high speeds for 20 seconds each. The slurry was passed through a screw press with a 0.02-inch (0.05 cm) screen. The juice was then clarified, acidified to pH 3.7 and heated to 80°c; it was refrigerated overnight at 1°c and the precipitate was filtered off.

The results were: Process (a). The press juice was dark green and appeared unacceptable for human consumption without excessive clean up. Also it was too low in protein (0.15%) for the clean up to be economically feasible. The low recovery of protein may be due to lack of sufficient crushing to release the protoplasm from a high proportion of cells. Also the pH of 6.6 did not appear to be conducive to maximum yield.

Process (b). Comminution, as opposed to chopping, improved the quantity of protein extracted from 23% to 37%. Two or three minutes' comminution were no more effective than one minute. The pH (9–12) of the extraction fluids was much higher than from chopped material.

Further work with hyacinth, containing 5.4% dry matter and 0.43% (fresh weight) crude protein, showed that a 1:3 ratio by weight of hyacinth to 0.05 N NaOH gave the best results. A yield of 49.4% of the crude protein of the whole plant was obtained.

Another test with water hyacinth having 5% dry matter and 0.235% (fresh weight) crude protein yielded 33.6% of the crude protein on extraction.

The authors comment that the yield of protein from the whole hyacinth plant is low compared with the extractions from legumes, reported by other workers, of 50–80%. It is possible that a greater proportion of the protein of the whole hyacinth plant is associated with fibrous structural material, rather then with chloroplasts as in leaves.

Giant water hyacinth from Guaiba river, Porto Alegre, Brazil

Water hyacinth on island beach, Rangoon, Burma

Previous Page Top of Page Next Page