The designations employed and the presentation of material in this publication do not imply the expression of any opinion whatsoever on the part of the Food and Agriculture Organization of the United Nations concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries.

CONTENTS

PHOTOGRAPHS.. xii

Acknowledgements.. xv

Chapter 1: Introduction.. 1

AQUATIC HABITATS.. 1

ECOLOGICAL CONSIDERATIONS.. 4

GLOBAL WARMING, FOSSIL FUEL AND NUTRIENT RECYCLE NEEDS.. 4

Chapter 2: The plant and its habitat.. 7

TAXONOMY.. 9

MORPHOLOGY AND ANATOMY.. 10

DISTRIBUTION.. 11

HISTORY OF DUCKWEED UTILIZATION.. 12

Chapter 3: Nutrient requirements of duckweed.. 13

INTRODUCTION.. 13

WATER TEMPERATURE. 14

WATER pH.. 14

MINERAL CONCENTRATIONS.. 14

WATER DEPTH.. 15

REQUIREMENTS FOR NPK AND OTHER MINERALS.. 16

Nitrogen requirements. 16

Phosphorus requirements. 21

Potassium requirements. 23

Sulphur requirements. 23

Sodium requirements. 23

Conclusions on mineral requirements. 27

Heavy metal accumulation by duckweeds. 27

MEETING MINERAL REQUIREMENTS.. 28

Fertilisers. 28

Manure. 29

Manure and biogas. 29

Miscellaneous systems. 32

Sewage. 32

RECORDED YIELDS OF DUCKWEEDS.. 33

Density of duckweed and yield. 35

Chapter 4: Integrated farming systems.. 37

WHY DO DEVELOPING COUNTRIES NEED TO EXAMINE THE POTENTIAL FOR INTEGRATED FARMING SYSTEMS? 37

RECYCLING OF NUTRIENTS.. 39

FIXATION OF N AND MOBILISATION OF P FROM PLANT GROWTH.. 39

INTEGRATED FARMING (THEORETICAL CONSIDERATIONS) 40

INTEGRATED SYSTEMS INVOLVING DUCKWEED PONDS.. 42

GROWING DUCKWEED IN AN INTEGRATED SYSTEM... 43

OTHER BENEFITS FROM INTEGRATING DUCKWEED INTO CROP FARMING.. 44

OTHER CONSIDERATIONS.. 44

CONCLUSIONS ON THE POTENTIAL OF DUCKWEED TO PRODUCE CHEMICALS OF IMPORTANCE TO HUMAN HEALTH 46

FARMING SYSTEMS FOR DUCKWEEDS.. 46

Why duckweed?. 46

RUSTIC METHOD FOR ESTABLISHING DUCKWEED.. 47

Photographs.. 49

Chapter 5: Duckweed as a source of nutrients for domestic animals 65

PRELIMINARY.. 65

CHEMICAL COMPOSITION OF DUCKWEED.. 66

Protein and amino acid composition. 66

DUCKWEED AS SOURCE OF NUTRIENTS IN AVIAN DIETS.. 67

Poultry - egg production. 67

Poultry - meat production. 71

Duckweed for family poultry. 71

Duckweed-fed carp poly-culture 84

The Mirzapur carp stocking strategy and carp growth rates 84

Feeding duckweed 86

Some ideas on duckweed use for carp production at small-holder farm level 87

Production of tilapia 90

Indirect use of duckweed to produce feed for fish 91

DUCKWEED AS A SOURCE OF NUTRIENTS FOR HUMANS.. 92

Safety considerations when duckweed enters the human food chain. 94

Chapter 6: Duckweed and its potential for waste management 95

INTRODUCTION.. 95

DUCKWEED TREATMENT AND SEWAGE WORKS.. 96

WATER RECYCLING AND HEALTH STANDARDS.. 99

Chapter 7: Overview... 101

Chapter 8: References.. 103

TABLES

Page

Table 1

The composition of duckweed harvested from a natural water source or grown on waters with minerals enriched

Table 2

Some mineral compositions of duckweed and their potential to remove minerals from water bodies

Table 3

Some examples of where duckweed might be used to cleanse wastewater of mineral pollution

Table 4

Field results of duckweed growth in near-optimal conditions

Table 5

Field results of duckweed growth in sub-optimal conditions

Table 6

Removal of nutrients by Lemna from a flow through pond fed by aerated pig waste

Table 7

Amino acid composition in aquatic plants (g/100g protein) grown on wash water from a pig farm in Cuba

Table 8

Composition of diets fed Topaz layers

Table 9

Performance of Topaz layers fed three isonotrogenous diets based on protein either from soya bean meal or duckweed after 2 weeks (Wolfia diet) or 10 weeks (Control and Lemna diets)

Table 10

Performance of poultry kept for egg production when dried Lemna powder replaced soya bean meal and some of the fish meal in the diets

Table 11

Mean intake of fresh foods by ducks

Table 12

Values for the composition of duckweed, broken rice and soya beans used in the studies of Men et al. (1995, 1996)

Table 13

Effects of replacing boiled soya beans with Lemna in a diet of sugar on intake and production of ducks.

Table 14

The effect of replacing “conventional” protein sources in a concentrate based diet for pigs with Lemna meal.

Table 15

Effects of replacing soya bean meal with Lemna on the growth of pigs over 3 months on a basal diet of molasses.

Table 16

Weight of different carp species in poly-culture fed with duckweed

Table 17

Potential monoculture of carp using duckweeds in an integrated system

Table 18

The effects of feeding tilapia increasing levels of Lemna

Table 19

The effects on the production of milk fish (Chanos chanos) of different fertiliser application

Table 20

Composition of protein extracts from three common aquatic weeds

Table 21

Some amino acids in a leaf protein extracted from aquatic weeds compared with leaf protein from alfalfa

FIGURES

Figure 1

Distribution of global aquaculture

Page

Figure 2

The changing pattern of world prawn production for human

Consumption

Figure 3

The influence of the concentration of N in culture water on crude

protein in duckweed (Spirodela spp) grown on diluted effluent

from a piggery

Figure 4

Relationship between root length and protein content in duckweed

Figure 5

The effect of N level in culture water on growth of duckweed

and its crude protein content

Figure 6

Yield and crude protein content of duckweed biomass growing on sewage waste water

Figure 7

The relationship between the quantity of P in duckweed and

the concentration of P in water

Figure 8

The uptake of SPK and N from sewage water held in galvanised

iron tanks

Figure 9

The effects of growing duckweed on saline mine waters with or without added N.P.K. fertiliser to optimum levels

Figure 10

Schematic representation of present duckweed framing in Vietnam

Figure 11

Diagrammatic representation of flow of nutrients through

simple biogas digesters to feed duckweed

Figure 12

Schematic diagram showing abattoir or intensive animal

production waste processing and biogas flow

Figure 13

The range of densities of duckweed biomass on the water surface after harvesting at which duckweed grows optimally.

Figure 14

Flow diagram showing the potential recycling of feed and faeces biomass from crop residues in an integrated farm

Figure 15

An example of an integrated farming system based on sugar cane and forage trees fractionated to provide feed for pigs and poultry

Figure 16

Variation in yolk colour of eggs from hens fed four conventional diets without artificial pigments with duckweed replacing soya bean meal

Figure 17

Eggs produced in 30 days by Local and Exotic (Tam Hoang) hens in semi-scavenging system with or without supplements of rice bran alone or plus duckweed (50:50)

Figure 18

Relationship between percent of diet DM consumed as duckweed and apparent DM digestibility in Mong Cai and Mong Cai/Large White crossed pigs

Figure 19

Effect of supplementing traditional (low-protein) diets with fresh duckweed for fattening pigs in Central Vietnam (Du Thanh Hang 1998)

Figure 20

Average weight of fish catch by month in Mirzapur duckweed-fed carp production

Figure 21

Average weight of fish catch after 13 months in Mirzapur

Figure 22

A schematic outline of a potential system for producing mono cultures of carp in a three pond system

PHOTOGRAPHS


				Page
Photo 1			The various species of Lemnaceae referred to in this publication (R A Leng)	49
Photo 2			Duckweed accumulation in the crocodile lagoon in Havana Zoo, Cuba (R A Leng)	49
Photo 3			Duckweed grown in a village for feeding to ducks (R A Leng)	50
Photo 4			Containers used for duckweed growth studies (R A Leng)	50
Photo 5			A young boy harvests duckweed in a village in Vietnam (R A Leng)	51
Photo 6			Duckweed mats fertilized with faeces introduced into the pond through a small basket in the middle of the pond (R A Leng)	51
Photo 7			A newly installed plastic biodigester in an ecological farm in Vietnam (Lylian Rodriguez)	52
Photo 8			Local (Mong Cai) sow eating duckweed in Vietnam (Lylian Rodriguez)	52
Photo 9			"Improved" hens (Tam Hoang breed from China) eating duckweed in locally-made bamboo cages in Vietnam (Lylian Rodriguez)	53
Photo 10			Scavenging hens and ducks supplemented with duckweed in Vietnam (Lylian Rodriguez)	53
Photo 11			Duckweed harvest is a daily routine for the small boys in a village in Vietnam (R A Leng)	54
Photo 12			Cassava waste and duckweed being mixed for duck feed in Vietnam (R A Leng)	54
Photo 13			Ducks being fed duckweed mixed with cassava meal in a village in Vietnam (R A Leng)	55
Photo 14			Ducks fed sugar cane juice and duckweed at the Ecological Farm of the University of Tropical Agriculture, Ho Chi Minh City, Vietnam (R A Leng)	55
Photo 15			Integration of pig fattening, a biodigester and duckweed pond in the Ecological Farm of the University of Tropical Agriculture, Ho Chi Minh City, Vietnam (Lylian Rodriguez)	56
Photo 16			Duckweed growing in ponds lined with polyethylene film (Lylian Rodriguez	56
Photo 17			Lining earth ponds with a cement-soil mixture prior to introducing duckweed (Lylian Rodriguez)	57
Photo 18			The finished duckweed pond lined with cement-soil (the previous arrangement using polyethylene film can be seen in the background)(Lylian Rodriguez)	57
Photo 19			Duckweed growing in the pond lined with a cement-soil mixture (Lylian Rodriguez)	58
Photo 20			Surrounding the duckweed ponds with forage cassva provides partial shade for the duckweed and an additional source of feed biomass for pigs and goats. Productivity of the combination of duckweed and cassava is close to 7 tonnes protein/ha/year (Lylian Rodriguez)	58
Photo 21			Contrasting quality of duckweed. One the left 35% protein duckweed from a pond with more than 20 mg N/litre in the water; on the right only 20% protein from a pond with less than 5 mgN/litre (Lylian Rodriguez)	59
Photo 22			Duckweed samples in glass jars prior to measuring root length. On the left grown on nutrient-rich water; on the right from water with few nutrients (Lylian Rodriguez)	59
Photo 23			Measuring root length to assess protein content; on the left high protein duckweed; on the right low-protein duckweed (Lylian Rodriguez)	60
Photo 24			The integrated system for a landless family. Palm leaf house, the latrine and 2 pigs provide excreta for the biodigester (with biogas reservoir in the roof) connected to a duckweed pond providing enough protein for 20 chickens (Lylian Rodriguez)	60
	Photo 25	Integrating the biodigester (covered with roof of palm leaves) with duckweed ponds and cassava planted around the biodigester and ponds (Nguyen Van Lai)			60
	Photo 26	Duckweed in ponds fertilized with effluent from a plastic biodigester and cassava for forage planted around the biodigester and ponds (Nguyen Van Lai)			61
	Photo 27	Duckweed growing in a village in Vietnam (Lylian Rodriguez)			62
	Photo 28	Duckweed growing in a pig farm in Vietnam (Lylian Rodriguez)			62
	Photo 29	Happy ducks harvesting duckweed! (Lylian Rodriguez)			63
	Photo 30	Duckweed being sold at a local market in Vietnam (Lylian Rodriguez)			63
	PPhoto 31	In Vietnam duckweed is a source of income for farmers [A bag of 20 kg is sold for 15,000 VND (1.10 USD)] as well as a source of protein for ducks and chickens (Lylian Rodriguez)			64
	Photo 32	A lady counting the profit from selling duckweed (Lylian Rodriguez)			64

Acknowledgements

My interest in Duckweed was stimulated by the work of the PRISM group in Bangladesh, and I wish to acknowledge their major contribution to the development of production systems using this remarkable plant. My friend and colleague Thomas [Reg] Preston encouraged me in my endeavours to promote research on the use of duckweed for small-holder farmers.

It was through the activities of the University of Tropical Agriculture, newly established on the campus of the University of Agriculture and Forestry, Ho Chi Minh City, Vietnam (now located at the Royal University of Agriculture, Phnom Penh, Cambodia), that many young scientists in developing countries have come to recognise the potential role of duckweed aquaculture as a component of sustainable farming systems for the future. Some of their contributions to development of duckweed production systems are to be found in this book.

I thank the Feed Resources Group of the Animal Production and Health Service of FAO, in particular Manuel Sanchez, for the opportunity to undertake this assignment. Special thanks go to Reg Preston for his thorough reading and editing of the final draft of the manuscript.

Chapter 3: Nutrient requirements of duckweed

INTRODUCTION

Like all photosynthetic organisms that utilise solar energy to synthesise biomass, duckweed only has requirements for minerals in order to grow. The plant has the capacity to utilise preformed organic materials, particularly sugars, and can grow without sunlight when provided with such energy substrates. In practice the capacity to use sugars in the medium as an energy source is irrelevant, as in most aquatic systems sugars are not present. However, they could be of some importance where there is a need to purify industrial effluents (e.g. waste water from the sugar industry or from starch processing).

Most research on nutrient requirements has focused on the need for nitrogen, phosphorus and potassium (NPK). However, like all plants, duckweed needs an array of trace elements and has well-developed mechanisms for concentrating these from dilute sources. From the experience of the NGO PRISM[1] in Bangladesh, it appears that providing trace minerals through the application of crude sea salt was sufficient to ensure good growth rates of duckweed in artificial ponds. However, considerable interest has been shown by scientists in the capacity of duckweed to concentrate copper, cobalt and cadmium from water resources where these have economic significance.

Mineral nutrients appear to be absorbed through all surfaces of the duckweed frond, however, absorption of trace elements is often centred on specific sites in the frond. The fertiliser requirement of duckweed depends on the source of the water. Rainwater collected in ponds may need a balanced application of NPK, which can be given as inorganic fertiliser or as rotting biomass, manure or polluted water from agriculture or industry. Effluents from housed animals are often adequate but may be too highly concentrated in some minerals, particularly ammonia. In such cases the effluents may need to be diluted to favour duckweed growth. Run-off water from agriculture is often high in P and N but the concentration may not always be appropriately balanced. Sewage waste water can be high or low in N depending on the pre-treatment system but is almost always adequate in K and P. Industrial waste water from sugar and alcohol industries for example are always low in N.

Little work has been done to find the best balance of nutrients to provide maximum growth of duckweed. Duckweed has evolved mechanisms that allow it to take up minerals preferentially and it can grow on very dilute media. The main variables that affect its growth under these circumstances are light incidence and water and air temperatures. The growth rate and chemical composition of duckweed depends heavily on the concentration of minerals in the water and also on their rate of replenishment, their balance, the pH and temperature of the water, the incidence of sunlight and perhaps day length. Its rate of production per unit of pond surface also depends on the amount of biomass present at any one time.

WATER TEMPERATURE

Duckweeds grow at water temperatures between 6 and 33°C. Growth rates increase with increase in water temperature, but there is an upper limit around 30°C when growth slows and at higher temperature ceases. In open lagoons in direct sunlight, duckweed is stressed by the high temperature created by irradiation and, in practice, yields are increased by mixing the cooler water in the lower layers of the pond with the warmer layers on the surface. This is done by agitating the water by splashing, which reduces the surface temperature of the duckweed matt.

WATER pH

Duckweed survives at pH between 5 and 9 but grows best over the range of 6.5-7.5. Efficient management should aim to maintain the pH between 6.5 and 7. In this pH range, ammonia is present largely as the ammonium ion, which is the most readily absorbed form of N. On the other hand, if the pH is too high the ammonia will be in solution as NH₃which can be toxic and can also be lost by volatilisation.

MINERAL CONCENTRATIONS

Duckweed appears to be able to concentrate many macro and micro minerals several hundred fold from water. On the other hand, high mineral levels can depress growth or eliminate duckweed, which grows best on fairly dilute mineral media. There is a mass of data on the uptake by duckweed of micro-elements which can be accumulated to toxic levels (for animal feed). However, their capacity to concentrate trace elements from a very dilute medium can be a major asset where duckweed is to be used as an animal feed supplement. Trace elements are often deficient in the major feeds available to the livestock of poor small-scale farmers. For example, in cattle fed mainly straw-based diets there are deficiencies of both macro and micro minerals.

Duckweeds needs many nutrients and minerals to support growth. Slowly decaying plant materials generally release sufficient trace minerals to provide for growth, which is often more affected by the concentrations of ammonia, phosphorus, potassium and sodium. There is an extensive literature on the requirements of duckweed for NPK. The capacity of the plant to acquire its requirements for micro nutrients from the aquatic medium is not usually considered to be a limitation to its growth. In the work by PRISM in Bangladesh, crude sea salt (added at 9 kg/ha of water surface area) was considered to be sufficient to provide all trace mineral requirements when duckweed growth rates were high at around 100 g of fresh plant material/m²/day.

WATER DEPTH

Depth of water required to grow duckweed under warm conditions is minimal but there is a major problem with shallow ponds in both cold and hot climates as the temperature can quickly move below or above the optimum for growth. However, to obtain a sufficiently high concentration of nutrients and to maintain low temperatures for prolonged optimal growth rate, a balance must be established between volume and surface area. Depth of water is also critical for harvest management, as a depth of more than 0.5 m can poses problems, particularly for resource-poor farmers. In contrast, if water purification is a major objective for cultivating duckweed, it is impractical to construct ponds shallower than about 2 m deep.

Incident sunlight and environmental temperatures are significant factors in determining the depth of water as undoubtedly duckweed is stressed by temperatures in excess of 30°C and below about 20°C growth rate is much reduced. In practice, the depth of water is probably set by the management needs rather than the pool of available nutrients and harvesting is adjusted according to changes of growth rate, climate changes and the nutrient flows into the system.

REQUIREMENTS FOR NPK AND OTHER MINERALS

Duckweed has evolved to take advantage of the minerals released by decaying organic materials in water, and also to use flushes of minerals in water as they occur when wetlands are flooded. Duckweed now appears to have the potential to be harnessed as a commercial crop for a number of purposes.

Water availability is likely to limit terrestrial crop production particularly of cereals in the coming years (see World Watch 1997). Water purification and re-use, particularly of water arising from sewage works, industrial processing and run-off from irrigation, appears to be mandatory in the future, both to reduce pollution of existing water bodies and to provide reusable water for many purposes including that required by people for drinking.

Nitrogen requirements

Duckweed appears to be able to use a number of nitrogenous compounds either on their own or through the activities of associated plant and animal species. The ammonium ion (NH₄⁺) appears to be the most useful N source and, depending on temperatures, duckweed continues to grow even at extremely low levels of N in the water. However, the level of ammonia N in the water affects the accretion of crude protein in the plant (see Table 1).

Table 1: The composition of duckweed harvested from a natural water source or grown on waters with enriched minerals (Leng et al. 1994)

	Crude protein	Fat	Fibre	Ash
Source	(% in DM)	(% in DM)	(% in DM)	(% in DM)
Natural lagoon	25-35	4.4	8-10	15
Enriched culture	45	4.0	9	14

The value of duckweed as a feed resource for domestic animals increases with increasing crude protein content. In studies at the University of New England, Armidale, Australia, the crude protein content of duckweed growing on diluted effluent from housed pigs increased with increased water levels of N from about 15% crude protein in dry matter with trace levels of N (1-4mg N/litre) to 37% at between 10 and 15mg N/litre. Above 60mg N/litre, a toxic effect was noticed perhaps due to high levels of free ammonia in the water. Whilst few experiments have been undertaken on the optimum level of ammonia required, these results give a guideline for the levels of N to be established and maintained in duckweed aquaculture to obtain a consistently high crude protein level in the dry matter.

Figure 3: The influence of the concentration of N in culture water on crude protein in duckweed (Spirodela spp) grown on diluted effluent from a piggery. The P levels in water varied from 1.2 to 6.1 mg P/litre (Leng et al., 1994).

In most practical situations the approach to growing duckweed is to find the dilution of water where N is not limiting growth and which supports high levels of crude protein in the plant. This is usually done by an arbitrary test. Serial dilution of the water source with relatively pure water (rain water) is carried out. Duckweed is seeded into each dilution and the weight change of the biomass recorded after, say, 4 weeks. In this way the appropriate N concentration is established.

A useful indicator of whether conditions in the pond are appropriate for growth of duckweed (Lemna spp) of high protein content is the length of the roots. Many experimental observations (Rodriguez and Preston 1996a; Nguyen Duc Anh et al 1997; Le Ha Chau 1998) have shown that over short growth periods there is a close negative relationship between root length and protein content of the duckweed. and with the N content of the water. Data taken from the experiment of Le Ha Chau (1998) are illustrated in Figure 4. In most small-scale farm situations it is not feasible to determine the protein content of the duckweed that is being used; nor can the nutrient content (especially nitrogen) of the water be estimated easily. To determine the root length of duckweed is a simple operation and requires neither equipment nor chemicals. By monitoring this characteristic, the user can have an indication of the nutritive value of the duckweed that is being produced and take the necessary corrective measures when the length of the roots exceeds about 10mm.

Figure 4: Relationship between root length and protein content in duckweed (Lemna minor) (Le Ha Chau 1998)

In duckweed aquaculture a source of N is essential and in many start-up systems, based on water effluent from sewage or housed animals, the project has been constrained by pre-treatments that denitrify the water and reduce ammonia concentrations. Most forms of aeration in sewage works are highly efficient in de-nitrification of waste waters, but this process compounds pollution problems. For instance where the effluent is high in P this promotes the growth of algae that fix N. In Australia the contamination of river systems with phosphorus often led to massive blooms of blue-green algae that are toxic to humans and animals.

Although there is an association of N-fixing cyanobacteria with duckweed, these are certainly not important from a standpoint of farming duckweed. Duong and Tiedje (1985) were able to demonstrate that duckweed from many sources had heterocystous cyanobacteria firmly attached to the lower epidermis of older leaves, inside the reproductive pockets and occasionally attached to the roots. They calculated that N fixation via these colonies could amount to 3.7 to 7.5 kg N per hectare of water surface in typical Lemna blooms, but that the association of cyanobacteria with Lemna trisulca was 10 times more effective.

Probably, under most practical situations, ammonia is the primary limiting nutrient for duckweed growth and the establishment of the optimum level for maximum growth of duckweed needs research, particularly in the variety of systems in which the plant may be expected to grow. The effect of time and lowering of N content of sewage water on yield and crude protein content of duckweed is shown in Figure 5 and Figure 6.

From recent research it appears that duckweed requires about 20-60mg N/litre to grow actively and from two studies [(those of Sutton & Ornes (1975) compared with those of Leng et al., (1994)] it is apparent that there is a complex relationship between the initial composition of the duckweed used in the research and the level of nutrients required. Stambolie and Leng (1993) showed that with duckweed harvested from a backwater of a river, and with an initial low crude protein content, it was only when the duckweed protein increased to the highest level that rapid growth of biomass commenced (i.e. at 3 weeks after introduction to the water) (Figure 6). By that time the N content of the water had declined to levels that were below the optimum that appears to be necessary for maximum protein levels (Figure 3).

In the work of Sutton and Ornes (1975), however, duckweed of a higher protein content was initially used and growth rate again peaked at about the third week (Figure 6) but by this time the crude protein content had declined to below 15%. These apparently conflicting results can be rationalised if there is a stress factor involved which requires a period of 3 weeks before it is overcome.

Figure 5: The effect of N level in culture water on growth of duckweed and its crude protein content. The experiments were conducted on duckweed collected from a river down-stream from a sewage works. The sewage water used in the incubation was taken from that flowing into the sewage works prior to the de-nitrification processing. The ponds were 2.5 m² (Stambolie & Leng, 1994)

(a) Crude Protein in duckweed

(b) Changing biomass on water

(c) Level of N in water

Growth rate appeared to be to be optimised at about 20mg ammonia N/litre, but to obtain maximum crude protein content required ammonia levels to be about 60 mg N/litre (see Figure 3). A further implication is that when the protein content of the duckweed is high at the commencement of the study, the duckweed can grow mainly through synthesis of carbohydrate. However, the variable results using duckweeds harvested from the wild and the slow "adaptation" to new conditions is obviously a confusing factor in interpreting any data of the requirements for duckweed for nutrients in such short-term studies.

Figure 6: Yield and crude protein content of duckweed biomass growing on sewage waste water (Sutton & Ornes 1975)

The most important issue is that duckweed increases its protein content according to the ammonia level in an otherwise adequate medium up to levels of 60 mg N/litre (see Figures 3, 5 & 6). For food or feed purposes there is a vast difference in the value of duckweed biomass depending on its protein content (see later). Rapid growth of duckweed is also associated with high protein accretion and low fibre content and fibre content increases where root growth occurs.

Phosphorus requirements

Duckweed appears to concentrate P up to about 1.5% of their dry weight and as such are able to grow on water rich in P, provided the N concentrations are maintained. The plant also appears to be able to draw on the pool of P in its biomass for its biochemical activities and once P had been accumulated duckweed will continue to grow on water devoid of P. On the other hand, the P in duckweed appears to be highly soluble and is released rapidly to the medium on death of the plant (Stambolie & Leng 1994).

The relationship between P in sewage water and in duckweed growing in this medium is shown in Figure 7. The concentration of P in duckweed growing on water with high P content was higher in our studies (Leng et al., 1994) than in those of Sutton and Ornes (1975). The time course of uptake of P by duckweed in static sewage water is shown in Figure 8. The differences in accumulation of P in the two studies may be explained by the relative rates of growth when the samples were taken. Duckweed has a very high capacity to concentrate P and levels of between 10 and 14 mg P/g dry weight have been achieved even when concentrations of P in the water have been as low as 1.0 mg/litre.

Figure 7: The relationship between the concentration of P in duckweed and the concentration of P in water. Filled squares are results from Sutton and Ornes (1975); the filled circles (upper values) are results from research in Australia (Stambolie & Leng 1994).

The important issue is that duckweed concentrates P when water levels are enriched with P and it appears to be readily available once the plant is disrupted or dies. The P level in duckweed is sufficiently high to be a valuable source of this nutrient for both plants and animals.

Potassium requirements

Vigorously growing duckweed is a highly efficient K sink, but only low concentrations of K in the water are needed to support good growth when other mineral requirements are satisfied. Most decaying plant materials would easily produce the K requirements of duckweed.

Sulphur requirements

Little work has been done to examine the S requirements of duckweed. The mechanisms for sulphate uptake have been studied since uptake of sulphate is the first step in the biosynthesis of S-amino acids. Such biosynthesis needs the integration of pathways providing carbon building blocks and reduced sulphur (Datko & Mudd, 1984). It is possible that S levels are at times limiting to growth or protein accretion because of the high level of S-amino acids in the plant when growth rate is high and ammonia in the medium is non-limiting. Sulphate salts appear to meet the requirements. As S is easily leached from soils it is unlikely to be deficient in farm-based production systems. The uptake of N, P, S and K by duckweed from sewage water is shown in Figure 7. The experimental design recommended for carrying out such studies is shown in Photo 4.

Sodium requirements

Sea salt (9 kg/ha/day) was applied as part of a fertiliser programme in pilot studies of duckweed farming in Bangladesh (see the discussion of PRISM's work in Chapter 6). These findings indicate that duckweed can accumulate sodium from low concentrations in the water. Using small galvanised iron tanks (see Photo 4) the effect of growing duckweed on saline waters from open-cast coal mines, with or without added nutrients, was studied. The growth rate and protein content of the duckweed is shown in Figure 8 together with the effects on mineral levels in the water. Duckweed grew on the water with or without added nutrients but the uptake of sodium was low.

Figure 8: The uptake of S, P, K and N by duckweed from sewage water held in galvanised iron tanks. Duckweed was harvested weekly by placing a piece of wood across the diameter of the tanks and removing half the duckweed (Stambolie & Leng, 1994).

The quality of the duckweed (indicated by its crude protein content) was maintained for some time by fertiliser application. The phosphorus requirement for growth was apparently low. This data is introduced here to show that even saline waters can be used to grow duckweed, although research is needed to investigate the needs for additional nutrients on saline waters.

Figure 9. The effects of growing duckweed on saline mine waters with (D) or without (o) added NPK fertiliser (Sell et al., 1993; Sell, 1993).

9(a) Crude protein of duckweed

9(b) Sodium content

Figure 9. The effects of growing duckweed on saline mine waters with (D)or without (o) added NPK fertiliser on dry matter harvested and P content (Sell et al., 1993; Sell, 1993).

9(c) Duck weed dry matter harvested: cumulative and residual pool

9 (d) Phosphorus content of mine waste water growing duckweed

Conclusions on mineral requirements

The capacity of duckweed to concentrate minerals permits it to grow under a wide range of conditions. Comparing water and duckweed dry matter levels for a range of minerals indicates the potential of duckweed to accumulate nutrients (Table 2).

Table 2: The mineral composition of duckweed and its potential to remove minerals from water bodies (calculated from the literature).

	Concentration in		Potential removal
	Culture medium	Duckweed tissue	at 10tonnes DM/ha/y
Element	(mg/litre)	(mg/kg DM)	(kg/ha/y)
N	0.75	60,000	600
P	0.33-3.0	5,000-14,000	56-140
K	100	40,000	400
Ca	360	10,000	100
Mg	72	6,000	60
Na	250	3,250	32
Fe	100	2,400	24

The advantage duckweed has for mineral removal must be weighed against the potentially detrimental effects of accumulation of heavy metals.

Heavy metal accumulation by duckweeds

All members of the duckweed family concentrate heavy metals: in particular cadmium, chromium and lead which may at times reach levels in the plant which are detrimental to both the health and growth of the plant in addition to creating problems where the plant is used in any part of a food chain eventually leading to human consumption.

The accumulation of heavy metals by duckweed is not normally a problem for those wishing to use duckweeds from natural water resources or effluent from human or intensive animal housing as the above metals are normally at extremely low concentrations in these resources

Duckweed is easily contaminated by heavy metals from industries such as tanning (chromium) and leachates from mining (e.g. cadmium). Great care is needed where water is contaminated to be sure that heavy metals do not get into the human food chain. On the other hand, duckweed may find a use in stripping heavy metals from industrial wastewater. Also their content of heavy metals can be used to indicate potential pollution levels of the water.

Cadmium appears to be absorbed by both living and dead duckweed plants and the cadmium is actively taken up by the plant Noraho & Gour 1996). Cadmium at high concentrations, that is the concentration that prevents vegetative reproduction (EC-50) was found to be 800ppb but duckweed grown in a medium of 2.2ppb still accumulated most of the cadmium from the incubation water over 7d. When the contaminated duckweed was fed to crayfish it accumulated in the muscles of the crustacean (Devi et al., 1996). It is therefore, extremely important to be sure of low cadmium levels in water prior to any large-scale use of duckweed as feed for domestic animals or humans.

Many reports are available on the uptake of metal ions by duckweed and the numerous interactions that occur. Duckweed will absorb and concentrate Cd, N, Cr, Zn, Sr, Co, Fe, Mn, Cu, Pb, Al and even Au. To define the rates of accumulation is not important here. It is enough to point out that when these minerals rise to higher levels than normal they may directly inhibit growth of the plant and any animal that consumes significant quantities of it. At low levels of accumulation, duckweed become a very useful source of trace minerals particularly for livestock and fish.

Problems of heavy metal contamination obviously arise where duckweed grows on industrial and mining waste. But in this case, the contaminating elements are known and therefore the problem should be apparent from the beginning of any study.

In conclusion, it is only where heavy metals are washed out in effluents from industry and mining that there is potential for duckweed to become toxic to livestock. In these situations, duckweed harvested from such sources should be disposed of in other ways (e.g. as a mulch for non-food crops such as trees).

MEETING MINERAL REQUIREMENTS

Fertilisers

A commercial balanced source of N, P, S and K, with sea salt to provide trace minerals, can be used in relatively unpolluted waters to meet the growth requirements of duckweeds. In the Mirzapur project (see chapter 6), muriate of potash (KCl), urea and super-phosphate (supplying P + S) were successfully used to produce duckweed on inundated land that also collected the effluent water from the local hospital. Fertilisers were applied on a daily basis, which together with the need for regular harvesting, had a high labour cost. Sea salt was added as a source of trace minerals.

Manure

Manure and other organic materials that decompose slowly are good sources of the nutrients required for growth of duckweeds. The skill here resides in how to control the nutrient inflow into the ponds. In many instances this can be established by trial and error. This has been apparently highly successfully in Vietnam where commercial producers of duckweed use ruminant and pig excreta. This is added to a small settling pond, the water from which then passes through a series of duckweed ponds before either entering the river or being used for irrigation. These systems appear to work because of long established experience with growing duckweed. All the ponds in the series apparently produce a good harvest of duckweed.

An extension of this method was seen in Bangladesh. The system was based on a simple toilet block, which may be just a hole in a concrete slab from which human excreta could be directed by gravity through a plastic pipe into a basket (usually split bamboo) situated at the centre of the pond and from which nutrients slowly diffused into the duckweed pond (see Photo 6).

Manure and biogas

The effluents from biodigesters, when suitably diluted, are very effective media for growing duckweed (Le Ha Chau 1998). These can be extremely simple systems, easily incorporated into a small farm situation. The biodigesters may vary in size and complexity from the low-cost plastic units (see Photo 7) through to industrial size biodigesters made of metal or concrete. In all cases, the excreta plus washings from animals held under penned conditions are collected, held in some form of settling pond to liquefy the solids and then pass into a closed container which allows anaerobic microbes to grow and convert the residual carbohydrates to carbon dioxide and methane. The gaseous effluent containing methane and carbon dioxide is collected and combusted for various purposes including household cooking. The water leaving the biodigester retains the minerals and with suitable dilution is a good media for duckweed ponds.

Figure 10: Schematic representation of present duckweed farming system in Vietnam.

Biodigester effluent from animal production is usually pH neutral and has a relatively high ammonia content. The mineral component of the diet effects the levels of nutrients in the water and therefore the need to dilute the effluent depends on the animal's diet. Ammonia-treated straw results in an effluent from cows that is high in ammonia. A simple system is shown in Figure 11. The system advocated by Dr Preston (Photo 8) is relatively simple to apply on small farms. A cow and calf, mainly fed crop residues, provide both urine and faeces, which suitably diluted with wash water are passed into a biodigester. The biodigester in this case is a simple polyethylene tube placed in a trench in the ground (Photo 7). The washings from the stall enter the biodigester and have a half-life of 10-15 days during which time up to 50% of the organic matter is converted to carbon dioxide and methane. The effluent is diluted and runs into narrow plastic lined channels or concrete channels in which the duckweed is seeded. The duckweed is harvested daily. The duckweed is then fed fresh as a protein supplement for pigs and poultry (Photo 9).

Figure 11: Diagrammatic representation of flow of nutrients through simple biodigester to feed duckweed.

Another system uses the effluent (washings) from a large number of animals housed under intensive management, or from abattoirs. The effluent is channelled through a lagoon covered with a heavy plastic film (5-10cm thick), and methane is collected from a convenient site beneath the plastic. The effluent is run into ponds, diluted and duckweed produced on the effluent (Figure 12). The heavy plastic is used as a safety precaution and thinner plastic can be used to cover the lagoons where more appropriate.

Figure 12: Schematic diagram showing processing of waste from an

abattoir or intensive animal production enterprise

Miscellaneous systems

Wherever there is a source of polluted water associated with industry or agriculture there is potential to purify it with duckweed. Each process, however, requires individual attention and it is beyond the scope of this book to make recommendation for all such systems. Some examples of situations where duckweed might have such an application are listed in Table 3.

Sewage

Possibly the best case can be put for the use of duckweed to remove P from human sewage which is mostly collected strategically at a point site in a township and, although treated to varying degrees, is often finally exported via rivers to the sea. There are now a number of commercially viable duckweed-based sewage systems that have been developed. These systems are expensive because of the need for high technology to ensure successful treatment of the sewage. It appears, however, that alternative types of plant, using chemical and microbiological treatments, are even more costly. On the other hand, an aquaculture component can often be added on to existing sewage purification plant. However, in small communities in the tropics the cultivation of duckweed in lagoons may be the only treatment necessary for sewage treatment.

Table 3: Some examples of where duckweed might be used to clean up waste water polluted with minerals and at the same time produce a feedstock of duckweed biomass

1. Effluents from:

· Dairies

· Piggeries

· Cattle and sheep feedlots

· Urban sewage

· Industrial waste from:

* Brewing & alcohol production (solubles)

* Milk processing

* Sugar factories

* Starch factories

* Wool scouring

* Abattoirs & tanneries

* Food processing

2. Run-off from:

· Agricultural practices

* Cotton growing

* Sugar industry

* Beef and sheep grazing industry

· Horticulture and nurseries

· Mining

* Sodium run-off water

* Heavy metals in other mining activities

· Parks and sporting facilities

RECORDED YIELDS OF DUCKWEEDS

The literature contains a great deal of information on the potential growth rates of duckweed. In many early studies the growth rates were measured under controlled conditions for short periods of time. In the absence of large scale field data obtained over 12 month periods these data have been used to estimate potential production rates. The yields estimated in this way need to be treated with some reservations as the data presented in Figures 4 and 5 point to serious problems in the interpretation of growth trials with duckweed grown under laboratory conditions.

The results in Table 4 are from research with close to optimum conditions for duckweed growth. Landolt and Kandeler (1987) concluded that under such conditions a production of 20g DM/m²/d is possible (equivalent to 73 tonnes dry matter/ha/year). Yields equivalent to180 tonnes DM/ha/year have been recorded. Under less than optimum conditions it is more realistic to target between 5 and 20 tonnes DM/ha/year (Table 5).

In practice the yields of duckweed often depend on the skill of the farmer in solving the problem of how to balance the mineral requirements of the duckweed and to identify with time the need for continuing and varying mineral supplementation. Water that is high in P and K and in trace elements need minimal but repeated inputs of a source of ammonia, so as to maintain N at around the 60mg N/litre level, at which point rates of growth and protein accretion are highest.

Yield of duckweed will depend on how the farmer monitors the duckweed system and whether a high protein final product is the objective. In the case of waters of industrial origin, where the aim is to reduce the P to a safe (low) level, continuous harvesting of duckweed and fertilization with urea are needed (as ammonia levels in the water decrease with continuous harvesting and become limiting to growth).

Table 4: Field results of duckweed growth in near-optimal conditions

Location	DM yield	Source
	(tonnes/ha/yr)
Louisiana USA	44-55	Mesteyer et al (1984)
Louisiana USA	27-38	Mesteyer et al (1984)
Louisiana USA	183	National Academy of Science (1976)
Southern USA	54	Said in Mbagwu and Adeniji (1988)
Southern USA	20	National Academy of Science (1976)
Israel	6-51	Oran et al (1987)
Israel	39	Heppher in Landolt et al (1987)

Table 5: Field results of duckweed growth in sub-optimal conditions.

Location	DM yield	Source
	(tonnes/ha/yr)
Thailand	10	Hassan and Edwards (1992)
Thailand	11	Hassan and Edwards (1992)
Thailand	10	Bhanthumnavin in Landolt et al (1987)
Israel	10-17	Porath et al( 1979)
Russia	7-8	Rejmankova in Landolt et al (1987)
Uzbekistan	7-15	Taubaev et al in Landolt et al (1987)
Germany	22	National Academy of Sciences (1976)
Germany	16	Schultz in Landolt et al (1987)
India	22	Rao et al in Landolt et al (1987)
Egypt	10	El Din in Landolt et al (1987)
Louisiana USA	9	Culley and Epps (1973)
Louisiana USA	20	Russoff et al (1980)
Florida USA	5-13	Reddy and DeBusk (1985)
Florida USA	17-21	Reddy and DeBusk (1985)
Florida USA	13	DeBusk et al in Landolt et al (1987)
Florida USA	19	Stanley et al in Landolt et al (1987)
Florida USA	23	Culley et al in Landolt et al (1987)
Florida USA	14-27	Meyers in Landolt et al (1987)
Florida USA	2-14	Sutton and Ornes (1977)

How often to feed urea into the system, and for how long duckweed growth can be maintained, can only be understood from research in the locality. Where the production of clean water is a major objective then it may be necessary to balance the supply of other nutrients as well as ammonia in order to end up with water that has had its total mineral composition decreased to levels that will allow its re-use.

In Table 6, the data obtained with wash water from pig production units show that, with skill, repeated fertilisation with urea followed by harvesting of duckweed could result in very low levels of P at the end of the process.

Density of duckweed and yield

The growth rate of duckweed under ideal light, temperature and pH would be exponential if there were no limitation in terms of mineral deficiencies or excesses. However, in practice many factors place a restriction on biomass yield. One of the most important is obviously plant density. The rate of harvesting duckweed is important since there is a minimum biomass at which yields will decrease and an upper biomass where yield will be limited by crowding, all other variables being equal. In a study where most of the conditions for growth were optimal the effect of harvesting indicated that when the duckweed density was above 1.2 kg/m^² or below 0.6 kg/m^²the growth rate decreased.

Table 6: Removal of nutrients by duckweed from a flow-through pond fed by aerated pig wastewater (Instituto de Investigaciones Porcinas, Havana, Cuba, unpublished observations)

	Concentration (mg/litre)
	Inflow water	Outflow water
COD	461	323
BOD	51	30
Total N	42	21
NH₃	17	2.2
Total P	6.4	3.3

Figure 13: Densities of duckweed on the water surface after harvesting at a rate at which duckweed grows optimally. Yield was 32 tonnes DM/ha/year. The upper density (filled squares) is when crowding limits growth and the lower density (unfilled squares) when growth is insufficient to prevent algal blooms (Stambolie 1994, from Leng et al 1994 )

Chapter 4: Integrated farming systems

WHY DO DEVELOPING COUNTRIES NEED TO EXAMINE THE POTENTIAL FOR INTEGRATED FARMING SYSTEMS?

Integrated farming systems, so long as they improve soil fertility (or at least maintain the same soil fertility), in the long term have major advantages which can improve the overall productivity of land without losing sustainability. The World Commission on Environment and Development defines sustainability as:

"ensuring that development meets the needs of the present without compromising the ability of future generations to meet their own needs".

However, development opportunities and aspirations change with changing economic considerations. Major increases in the cost of food production are likely to arise as the cost of fuel (fossil) increases relative to income. Fuel prices will surely increase in the future in response to:

· depletion of world reserves (Fleay, 1996)

· economic decisions taken at government level to reduce fuel use because of global warming.

· economic downturns which put pressure on gasoline prices.

The cost of food production in a country is highly dependent on fuel prices as was shown in the 1998 Asian financial crisis. The two cost factors that will be affected most in agriculture are: mechanisation, where fuel is directly consumed in crop farming; and fertiliser availability and application since the cost of chemical fertilizers is highly related to the price of fossil fuel. The fuel crisis in Cuba, brought about by removal of economic support from Russia and the embargo by the United States, has brought about a return to animal traction in the past 6 years and a massive decline in crop yields through decreased use of inorganic fertilisers.

Farms that produce for the market export mineral nutrients in the products that are sold. There are also losses of minerals in the effluents that arise from intensive crop and livestock activities. In order to maintain food production, these minerals must be replaced. In industrialised farming systems this is done largely by inorganic fertilisers produced and delivered at an increasing cost measured in terms of fossil fuel.

High level use of NPK has resulted in the sustained high yields of crops in industrialised countries and, in the last 20 years, has contributed to greatly improved yields in developing countries where these inputs have been used together with crop varieties of greater yield potential.

The other major issue in crop production has been the increased use of water, the sources of some of which are irreplaceable. Levels of application of fertiliser and water are almost always in excess of plant needs. Run-off water contaminated with minerals has created great problems of salination and eutrophication in river and pond systems throughout the world, changing the aquatic ecology of whole regions. Integrated systems aim to minimise (or prevent) loss of nutrients from a farming system and, in many situations, to conserve water for reuse (Preston & Murgueitio, 1992). Integrated farming systems require considerable skills in operation in order for them to be economic, high-yielding and sustainable. Integration may be developed on a single land holding or may be more easily applied where a number of farms combine their activities to develop an integrated system where the minerals leached from the land by farming are returned to the land via good conservation practices involving a number of farms. In Australia the "Land-Care Movement" involves usually a number of land-holders who combine their efforts to conserve a whole watershed. Similarly in India a "watershed" approach to sustainability has been implemented through ICRASAT. Integrated farming systems have the potential:

· to prevent land degradation

· to minimise external (costly) inputs

· to conserve resources otherwise lost through effluents

· to increase the income and standard of living of the farmer

· to maintain the fertility of the land.

Integrated farming systems to be sustainable require that:

· losses of minerals within the system are minimised and/or eliminated (this could be termed nutrient recapture).

· minerals, exported in products, are replaced by sources generated on the farm or from by-products of agro-industries (e.g. minerals in water from industries such as sugar production, fertiliser production or from industrial scale biodigesters).

RECYCLING OF NUTRIENTS

Integrated systems were traditional in most developing countries prior to the "green revolution" and many ancient societies recognised and put into practice sustainable cropping systems often through application of taboos against practices that caused degeneration of resources necessary for food security. In more recent times (about the 1920's), this took the form of integration of crop and animal production. The animals were an intermediate element in conversion of crop residues and other wastes to manure, which was then returned to the land. In some countries, composting and biodigesters were instrumental in recycling nutrients within the farm(s). The net result was that the minerals in the biomass produced on the farm were recycled by re-incorporating them in the land via human or animal excrement. Often the animal was a draught animal - now replaced by tractors in the industrial countries. However, even in these systems mineral loss was considerable in water runoff and in products sold off from the farm.

FIXATION OF N AND MOBILISATION OF P FROM PLANT GROWTH

Within the integrated farming system, there must be strategies to encourage N fixation and to increase P availability. Accumulation or at least the maintenance of N levels in soils through N-fixing plants (e.g. legumes), and the extraction of minerals from effluents by aquatic plants, are strategies that were used by small-scale farmers only a few decades ago. For example, in Kashmir, aquatic plants growing on the bottom of lakes are harvested for use as fertilisers. Seaweeds have been used for application to soils wherever they were washed ashore. Ruminants can harvest considerable quantities of N and P when they graze non-cultivatable land and these minerals can then find their way into crops via manure.

The major constraint to the establishment of integrated farming systems is the level of management that must be exerted, and this may be a major factor in the case of the small-holder farmer. An exception to this statement, however, was seen in Vietnam and Bangladesh where the collection of all animal and human wastes into ponds and the subsequent growth of duckweeds has proved to be relatively free of problems and is very skilfully managed by a number of cooperating farmers. However, if these tropical systems had to be managed for production of quality water, as well as feed for ducks and fish, a greater degree of control of the duckweed growth would need to be exerted.

INTEGRATED FARMING (THEORETICAL CONSIDERATIONS)

One of the major reasons for promoting a role for livestock in cropping systems in developing countries is to utilise crop residues efficiently, thus eliminating waste and optimising the use of the total biomass produced within the farm. The small-holder farmer has often a priority requirement for draught power with animal products (milk and meat) as secondary considerations.

Integrated crop and livestock production systems can be highly efficient in utilization of natural resources. Crop residues are used as livestock feed, the waste products (e.g. faeces and urine) are fed into a biodigester and the effluent used to fertilise ponds for aquatic plant/algae production, with fish farming as the terminal activity. These systems are worthwhile pursuing as a means of providing nutrients / fuel for the family, reducing the burning of firewood, minimising fossil fuel combustion and reducing environmental pollution (Preston, 1990).

The array of integrated strategies that could be developed is large. They all have as a central core a flow of nutrients through a number of systems. At each of these steps research can be brought to bear to optimise the partitioning of the available biomass into food, fuel and residues (see Figure 14). The environmental attributes of such systems are that methane emissions into the atmosphere and fuel (fossil fuel and firewood) use are minimised. In addition the efficient and also total harnessing of the energy from high producing crops reduce the land areas required per unit of product (see Preston, 1990).

Figure 14: Flow diagram showing the potential recycling of feed and faeces from biomass from crop residues in an integrated farm.

A complete discussion of these systems is beyond the scope of this document but two examples are:-

· The use of aquatic plants / algae grown on biodigester effluent for protein production for pigs, poultry, ruminants, rabbits and horses particularly in the humid tropics and

· The farming of duckweed on biodigester effluents to feed fish.

Figure 15: An example of an integrated farming system based on sugar cane and forage trees fractionated to provide feed for pigs and poultry (the juice and tree leaves), sheep (the cane tops and tree leaves), fuel for the family (bagasse and firewood) and litter for sheep and earthworms (bagasse), with recycling of excreta through biodigesters to provide fuel (biogas) and fertiliser (the effluent) for water plants in ponds and for the crops (Preston, 1990)

INTEGRATED SYSTEMS INVOLVING DUCKWEED PONDS

Considerable interest in the use of duckweed was stimulated by the publication of a booklet (Skillicorn et al. 1993) on duckweed aquaculture based on the experiences of a project operated by PRISM in Bangladesh. In Cuba, major research activities have been developed as a result of the pioneering work of the PRISM group. It is of significance that both these countries have relatively high-priced fuel and, in Bangladesh, draught animal power is still the main source of farm power. In Cuba, the farmers have reverted to draught animal power mainly because of the increased price of diesel and gasoline fuel. Neither country, however, has been able to establish efficient integrated farms. In both countries the objective behind the research on duckweed has been to provide food for carp and/or tilapia production with some spin-off for pigs and poultry.

A system that appears ready to be put straight into farm practice arises from the work of Dr. Preston and his colleagues in Vietnam. It incorporates a duckweed production system into a rice farm or market garden. It depends on a typical 1 to 0.5 ha farm with a range of livestock species (e.g. one milking animal, a calf and a bullock for work and with additional small ruminants [goats], pigs and hens).

The pig is a crucial animal in many of the farming systems of small-holder farmers in Vietnam as the manure is used to maintain rice yields. The pigs are fed on the by-products of the household with vegetables as the only supplement. The recycling of the excreta of these animals has been the mechanism by which soil fertility has been maintained over the centuries in much of the Mekong Delta. However, these systems are becoming less integrated because of the impact of imported technologies (2-wheel tractors replacing buffaloes) and introduction of industrial country standards ("lean" pig production) with imported "balanced" feeds.

GROWING DUCKWEED IN AN INTEGRATED SYSTEM

The nutrients required to support high rates of duckweed growth have been discussed. In practice, however, "standard" requirements only provide a basis for the "adviser" to give recommendations to the farmer. Most scientists are distracted by the felt need to establish standards of "nutrient requirements". The application of precise levels of nutrients into any system is difficult, particularly where these have to be met from the farm resources. However, mineral analysis of the water in a research laboratory may help to identify a critical limitation in a duckweed production system under farm conditions.

Duckweed production should be based on the effluents arising from plant and animal activities. Often the effluent from plant production (drainage) is too low in nutrients for high growth rates of duckweed. An exception to this occurs in some areas where high levels of fertiliser are applied to crops under irrigation. In Pakistan close to Faisalabad, duckweed growth in drainage ditches can sometimes be so great as to create major problems by blocking the irrigation pumps. Similarly, the run-off from cotton production often provides a good medium for duckweed growth. On the other hand, effluent from intensive animal production almost always is too concentrated a source of minerals particularly ammonia, and needs dilution with other sources of wastewater.

Levels of ammonia-N and phosphorus in effluents and wastewater can be measured in the laboratory and the results used as the basis of recommendations for the appropriate dilution of pond water for duckweed farming. On the other hand, chemical analysis of water is not feasible or affordable for large numbers of resource-poor farmers. However, some practical recommendations based on simple research at the farm level can be given to assist farmers to establish a duckweed operation.

OTHER BENEFITS FROM INTEGRATING DUCKWEED INTO CROP FARMING

In aquaculture systems, duckweed has been largely promoted as an opportunity crop for use as animal or fish feed. This is particularly appropriate where ponds have become unusable for other purposes because of pollution, emanating from fertiliser run-off or from washout of animal / human excreta. Such ponds are abundant in countries such as Bangladesh, where there are an estimated 1.3 million ponds (average size 0.11 ha) covering 147,000 ha. Only 46% of the ponds contain fish. The ponds have multiple uses: for bathing (washing), irrigation and for watering livestock. These activities interfere with fish culture but could make them useful for duckweed production. A recent World Bank review suggested that about 40,000 ha of ponds could be brought into duckweed production with potential to produce annually from 160,000 to 800,000 tonnes of duckweed (assuming a range of from 4 to 20 tonnes DM/ha/year) which could be made available to poultry farmers. The potential value of this can be seen from the fact that the higher quantity of duckweed represents twice the availability of 'home grown' concentrate feed in Bangladesh.

The major problems of developing such a system of production are associated with multiple ownership of the ponds, lack of credit and the unavailability of extension services. Another constraint is the lack of a marketing mechanism, including quality control, that is necessary if locally produced duckweed is to compete with imported protein meals.

OTHER CONSIDERATIONS

A major benefit of using duckweeds is emerging. There is accumulating evidence that duckweeds release compounds that have insecticidal properties that act against the larval stages of mosquitoes. Thus the development of duckweed aquaculture may have implications for mosquito control in rural areas in the wet tropics where malaria is again becoming a serious problem.

Eid et al. (1992a) published evidence to show that an extract of Lemna minor had insecticidal action against the mosquito Culex pipens pipens. The same extract contained synomones, which also repelled oviposition by the female mosquito. Where sub-lethal doses of synomones were added to water it was found that all larval stages of the mosquitoes were malformed. Duckweed synomones added to water also repelled the ovipositing of Piophila casei, and affected larval development and reduced survival. Similarly Spodephera littarolis larvae were malformed when synomones from Lemna minor were added to their culture medium.

If the insecticidal properties of Lemna minor and other species of duckweed are sufficient to control mosquito populations it could have a major beneficial impact on the health of people in areas where mosquito-borne diseases are endemic and resistance of the parasitic stage to drugs has increased. It is also a further inducement to cultivate duckweed widely for water treatment (purification) and animal feed. A further potential is the commercial cultivation of duckweed as a source of insecticides in water where it is difficult to spray for control of mosquito larvae or where the use of other control measure is impractical.

Other research workers have also associated the presence of duckweeds with reduced (or elimination) of mosquito development. For example, Marten et al. (1996) showed that Anopheles albimanus populations were negatively correlated with the amount of cover of the water by Lemna. The relative cover of water surface with duckweed was also negatively correlated with populations of fish and other insects indicating how intricate the associations are in natural ecosystems.

Eid et al. (1992b) have made the observation that the mosquito Culex pipens pipiens never colonised sewage water covered with duckweed and Bellini et al. (1994) observed that Lemna covering the surface of rice paddy-fields strongly effected mosquito populations. Lemna trisulca appears to produce allelo chemicals that are active against algae (Crombie & Heavers 1994) and Mesmar and Abussaud (1991) suggested that extracts of Lemna minor were active in inhibiting the growth of Staphylococcus aureus. The role of duckweeds in preventing algal growth can be by shading, by perhaps the production of algacides if this can be proved, and in addition because they lower the nutrient supply, particularly of P. Cholera has long been associated with seasonal coastal algal blooms off Bangladesh. Fluorescent antibody techniques have shown that a viable, non-cultivatable form of Vibro cholerae exists in a wide range of marine life, including algae. In unfavourable conditions, V. cholerae assumes spore-like forms, which as conditions improve revert to a readily transmittable and infectious state. Algal blooms, which are associated with eutrophication, have been related to the spread and persistence of cholera. Prevention of algal blooms may therefore be of considerable benefit (see Epstein, 1993).

CONCLUSIONS ON THE POTENTIAL OF DUCKWEED TO PRODUCE CHEMICALS OF IMPORTANCE TO HUMAN HEALTH

Although the literature is sparse and not totally convincing on the potential of duckweed to control mosquito populations, the author has heard farmers in many countries express opinions that growing duckweed on ponds does help to control mosquitos. Farmers in a group of villages in Vietnam that traditionally produce duckweeds were adamant that mosquitoes were not a problem so long as their lagoons were covered with duckweed.

Research into the insecticidal properties of duckweed is worthy of follow-up. If the production of natural insecticides can be promoted, and to this role is added water purification and providing a natural food resource for animal production, there could be far-reaching implications for duckweed production. A new naturally-occurring insecticide, produced from duckweeds, could be as revolutionary as the discovery of pyrethrins.

FARMING SYSTEMS FOR DUCKWEEDS

Why duckweed?

A number of aquatic plants including duckweed have great potential for development for various purposes. Aquatic plants may or will be grown in developing countries where:

· there is an unused area of standing water available that is either free or is relatively inexpensive to purchase, rent or lease.

· fresh water fish / crustacean production is impractical, not practiced or there are other constraints to their production, such as pollution.

· there is a need to clean water of chemicals before reuse or release into the aquatic ecosystem of rivers / deltas or seas.

· there is a market for the product or the product can be integrated into a system of production enhancing the economic viability of the farm either being used as mulch, fertiliser, feed/food and perhaps even fuel.

· there is a levy on industries in disposing of water contaminated with chemicals.

· legislation is enacted in order to clean up vast areas of ponds or wetlands that have become unusable for, in particular, fish production.

· duckweed mats on standing water are shown to reduce the health hazards

Candidates for use in any of these applications include duckweed, Azolla, Pistia, Eichhornia and a few lesser known aquatic plants. There are major advantages for floating aquatic plants as the water depth is not critical and harvesting does not necessarily disturb the underlying ecosystem in the bottom of the pond. Ease of harvest is important and Azolla and duckweed are readily harvested, but have the disadvantage of having to be protected from wind and water currents to encourage total coverage of lagoons and hence to maximum yields. There is some evidence for a symbiotic association of Lemna and N-fixing bacteria, but on water with a low N content Lemna growth is slow and the product is low in protein. On water high in P and low in N, it is more appropriate to grow Azolla with its highly productive association with N-fixing bacteria. However, Azolla is more difficult to manage in a continuous production system, because of susceptibility to insect damage. Addition of N fertiliser in aquatic media, or the presence of high levels of N derived from livestock wastes, mostly removes the major advantage of Azolla, which is its ability to grow on low-N water.

RUSTIC METHOD FOR ESTABLISHING DUCKWEED

A simple approach to establishing a duckweed pond system is often the best way to start.

This example assumes the availability of wastewater emanating from an animal production unit. First, the wstewater has to be collected in some suitable settling pond. To obtain information on the water's potential to grow duckweed, the water is serially diluted in small containers with water relatively free of minerals. Duckweed is seeded into each container and its relative growth monitored by eye. It quickly becomes apparent what is the appropriate dilution. This can be further refined by successive harvesting from the containers to determine the dilution at which duckweed grows at the greatest rate. This system can be recommended where the objective is to produce duckweed and there is no constraint to disposing of the effluent, nor a need to recycle the water.

In general, it appears that in most systems N quickly becomes the limiting nutrient as duckweed mats grow. The second potentially limiting nutrient is P. Thus there are sometimes great benefits in providing extra N (as urea) at the end of the growth period following harvest of half the duckweed mat. This allows further growth of duckweed and further reduction of P content in the effluent water. Water can be cleaned effectively by growing and harvesting duckweed only when there is a well-designed succession of fertiliser applications that rebalance NPK for growth after each harvest. Eventually the minerals may be reduced to acceptable levels.

Chapter 5: Duckweed as a source of nutrients for domestic animals

PRELIMINARY

It has been shown that farmers, particularly in South East Asia and probably elsewhere, have developed methods for the use of duckweed as a source of nutrients for livestock. However, there has not been the controlled experimentation that has been typically used to develop livestock feeding systems based on such commercial crops as soya beans and maize. There are, however, a number of reports in the literature on the use of duckweeds as feed supplements for fish and livestock. These report research with domestic animals in which normal feed protein sources have been replaced by duckweed meal on an iso-nitrogenous basis in complete diets based on compounded concentrate feed.

Duckweed is highly variable in composition. It grows slowly on low nutrient waters and in such situations is high in fibre, ash and carbohydrates but low in crude protein. In contrast, when grown on water high in ammonia and other minerals it grows rapidly and has a high protein content associated with a high ash and is often lower in fibre. Because duckweed responds quickly to the availability of nutrients it often has highly variable levels of some nutrients, which makes it difficult to prescribe the amounts needed for livestock and fish over an extended period. Careful interpretation of some of the reported data is required when the quality of duckweed that was given to the livestock was not consistent throughout the study. In terms of domestic animal / fish nutrition, duckweed may be used in many ways. These include:

· As a total feed

· As a supplemental source of:

- protein

- phosphorus and other major minerals

- trace minerals

- pigment for egg yolk / flesh of chickens

- vitamin A and the B group

- fibre in low-fibre diets for pigs and poultry

Duckweeds have been largely researched as a total feed for fish, including carp and tilapia production, as a protein supplement for pigs and poultry (including ducks) and as a source of fermentable N and minerals for ruminants.

The research on duckweed as a feed is summarised below. The uncertainty of the conclusions, and the difficulty in making clear recommendations, largely stems from the fact that the quality of the duckweed used (i.e. its nutrient density) was variable. However, as a resource that can be harvested for labour costs alone in natural conditions, it obviously represents a valuable asset to the resource-poor farmer. In many countries it could have a low cost where it is grown on sewage and it has to be disposed of from the treatment plant at a subsidised price. It appears to be a resource that is most conveniently used by the small-holder farmer, particularly in an integrated farming system. Unfortunately, much of the research has attempted to demonstrate the value of duckweed as a protein source in diets that are most commonly used in industrial production systems. This is particularly true of the research with poultry and yet its major application probably lies in the more difficult situation of increasing animal production on small farms.

CHEMICAL COMPOSITION OF DUCKWEED

Protein and amino acid composition

The crude protein content of duckweed depends on the N content of the water upon which it grows. Some publications have indicated that there are variations in amino acid content of duckweed protein. High levels of lysine have been reported from the duckweed research programme in Bangladesh. However, it appears that the protein component of most aquatic plants, including duckweed, have similar amino composition to terrestrial plant proteins (Table 7). In this respect, the amino acid composition is influenced by the major enzyme protein in plants, which is ribose bisphosphate carboxylase. Protein extracted from Lemna minor when fed to rats had a similar nutritive value as a wheat flour diet. This indicates that Lemna meal has a relatively high biological value for rat growth (Dewanji & Matai, 1996).

Table 7: Amino acid composition in aquatic plants (g/100 g protein) grown on wash water from a pig farm in Cuba (unpublished data from Instituto de Investigaciones Porcinas, Havana). The wash water was collected and aerated to reduce total N (Figueroa, Vilda, personal communication)

	*Azolla*	*Duckweed*	Water hyacinth	Soya bean
Crude protein	31.0	28.0	19.0	44.0
Lysine	3.4	3.7	3.1	6.6
Histidine	1.7	1.7	1.4	2.5
Arginine	4.6	5.1	3.7	7.3
Aspartate	-	-	-	-
Threonine	3.5	4.2	3.3	3.9
Serine	-	-	-	-
Valine	5.1	5.8	4.5	4.6
Methionine	1.4	1.5	1.2	1.2
Isoleucine	3.8	4.3	3.1	4.5
Leucine	7.1	7.8	5.6	7.7
Tryptophane	3.5	4.2	3.3	3.6

In the above study, the duckweed protein had a low lysine content considerably below that of soya bean protein, which is contrary to the data produced by Rusoff et al (1980), Skillicorn et al. (1993) and Dewanji (1993) . The balance of evidence from the feeding trials (see later section) supports the conclusions of the latter three groups of researchers which is that duckweed protein has as good an array of essential amino acids as has soya bean.

DUCKWEED AS SOURCE OF NUTRIENTS IN AVIAN DIETS

Poultry - egg production

The value of duckweed as a protein supplement for poultry was recognised some time ago (Lautner & Muller, 1954; Muzaffarov, 1968; Abdulayef, 1969). Truax et al. (1972) showed that dried duckweed was superior in protein quality for poultry as compared to alfalfa meal and could fully substitute for alfalfa meal at 5% of the total diet.

Interpretation and extrapolation from much of the earlier studies are confusing because the duckweed was harvested from natural sources and often the material could have been "old" and low in protein but high in fibre. This situation led Haustein et al. (1990) to reassess the value of duckweed as a protein supplement for pigs and poultry. They established studies to examine the potential to substitute not only alfalfa meal, but also fish meal and / or soya bean meal with duckweed meal in a compounded feed.

The diets used by these researchers at the University of La Molina, Lima, Peru were based on those used in the intensive egg production industry (Table 8). Duckweed was harvested from a tertiary sewage effluent and lagoon run-off and in general had a medium level of protein (33% CP in DM). Both Wolfia and Lemna species were harvested and their estimated metabolisable energy level was 1,200kcal/kg (in young broilers) and 2,000kcal/kg (in mature cockerels) indicating a poor overall digestibility of duckweed for monogastric animals (see also Hanczakowski et al. 1995). It might be inferred here that there was considerable intestinally indigestible carbohydrate (fibre?) present.

Table 8: Composition of diets fed to Topaz layers (Haustein et al. 1990)

Ingredient (%)	Diet
	Control	*Lemna* (15%)	*Wolfia* (15%)	*Lemna* (25%)	*Lemna* (40%)
Ground corn	52	51	51	48	50
Wheat middlings	19	16	16	*	*
Fish meal (65% CP)	7.5	7.5	7.5	2	-
Soya bean meal (46% CP)	11.0	-	-	-	-
Duckweed	-	15	15	25	40
Fish oil	2.5	2.7	2.5		2.5
Minerals and vitamins	+	+	+		+
Calculated ME (kcal/g)	2,800	2,800	2,800	2,840	2,800
Crude protein (%)	17	17	17.5		17.0

* Approximate as the level of wheat middlings was not stated by the authors and small amounts of other carbohydrates were included in the diet to balance the energy.

These studies compared dehydrated duckweed (prepared in meal form), as complete replacement for soya bean meal, and all or the greater part, of the fish meal, in commercial diets used for intensive production (Table 8). The results indicate that both species of duckweed are as good as soya bean meal as a source of essential amino acids as there were virtually no differences in egg production between groups of birds on the different treatments (Table 9).

Table 9: Performance of Topaz layers fed three iso-nitrogenous diets based on protein either from soya bean or duckweed after 2 weeks (Wolfia diet) or 10 weeks (Control and Lemna diets) (Haustein et al. 1990)

	Diet
	Control	*Lemna*	*Wolfia*
		(15%)	(15%)
Egg Production * (%)	89	90	90
Eggs per week	6.2	6.3	6.3
Feed conversion efficiency (g/g)**	2.3	2.4	2.4

* As a percentage of the egg production when all birds were fed the control diet in the pre-experimental 2 week period

** total egg produced (g) divided by total feed consumed.

Table 10: Performance of poultry kept for egg production when dried Lemna meal replaced soya bean and some of the fish meal in the diets shown in Table 8. The experiment lasted 18 weeks.

	Diet
	Control	*Lemna*	*Lemna*
		(25%)	40%) **
Egg Production (eggs/week)	5.9	5.9	5.5
Feed consumed (g/d)	131	131	125
Feed conversion efficiency (g/g)*	2.4	2.5	2.5
Liveweight change (g/18 weeks)	46	114	-118

* total egg weight produced divided by total feed consumed

**Lemna completely replaced both fish meal and soya bean meal in the diet.

Even when Lemna meal was increased to 40% of the diet, egg laying was sustained for 18 weeks but the birds were losing weight and therefore it could be anticipated that eventually egg production would have decreased (Table 10). These studies clearly demonstrated the value of a duckweed meal, that was of only medium quality (33% protein in dry matter), as a source of essential amino acids for egg production. The encouraging outcome from these experiments led to a number of other issues being researched and the following observations were made:

· There was no contamination with faecal organisms of the meat from birds consuming duckweed produced on sewage water.

· The quality of the eggs was probably not changed, however, the authors do suggest an improved taste and preferred higher pigmentation of the egg yolks

· There were no problems of heavy metal concentrations in the duckweed from sewage farms.

· Bulky, wet faeces were produced by the birds given 40% Lemna in the diet. This has implication for the large commercial producer but is irrelevant to the small farmer and may even be an advantage.

This demonstrated quite effectively that in areas where protein resources are scarce, duckweed represents potentially a high quality protein source that can be safely exploited for poultry production. Recent research in Australia investigated egg production and egg characteristics in two strains of layers Tegel Hi-Sex and Tegal Super Brown birds (Nolan et al., 1997). The birds were changed from a conventional layer diet to diets in which duckweed (Spirodela) represented 10, 30, 50, 80, 120 and 200g/kg of feed replacing both grain and soya bean meal to retain a diet with 160g crude protein, 40 g Ca, 10 g P and 11.3 MJ of ME per kg feed. Egg mass production was slightly reduced in the Hi-Sex hens given diets containing the higher duckweed content but it was more markedly reduced in Super Brown hens. Duckweed in the diet increased the pigmentation substantially (Figure 16).

Figure 16: Variation in yolk colour of eggs from hens fed four conventional diets without artificial pigments but with duckweed replacing soya bean meal.

These more recent results confirm the value of duckweed as a source of protein, minerals and pigment for poultry. The researchers in these particular studies were motivated by the potential for large-scale production of duckweed on, for instance, sewage lagoons. The costs associated with large-scale production, which involves modification of sewage systems, implementation of mechanised harvesting, drying and processing, are not known. In any event such a system has no application to small-scale farmers who produce eggs and meat from birds that scavenge for a proportion of their feed.

Poultry - meat production

Recent studies have demonstrated that on conventional diets for young broiler chickens, replacing a conventional protein source with duckweed meal retarded growth as levels increased (Haustein et al., 1992b, 1994). Older broiler birds had excellent growth characteristics when fed relatively high levels of duckweed meal. This is of significance to the factory production systems for poultry where margins per bird are often small and slight decreases in profitability per bird are important.
As with layers, the results have no significance for the small-scale farmer, who would almost certainly feed the duckweed fresh and whose birds would be in a semi-scavenging system.

A major question arising from this work is what would have been the value of the duckweed as a protein supplement had the duckweed been fertilised with a little extra nitrogen (urea) so that the protein level had approached the upper limit (about 40% CP)?

Emphasis now needs to go to the other end of the spectrum of production systems for poultry. Major research efforts are needed to find ways by which duckweed can increase egg and meat production from non-conventional diets as used by small-holder farmers at the village level.

Duckweed for family poultry

Poultry have a well-developed capacity to select a balanced diet from individual resources made available to them (Mastika & Cumming 1985). Local chickens in a semi-scavenging system in Cambodia were fed broken rice alone (control treatment) or broken rice and either fresh duckweed or ground whole soya beans, offered in separate feeders (Hong Samnang 1998). Supplementation with either duckweed or ground soya beans increased growth rate compared with control birds but the economics strongly favoured the duckweed treatment. Observations in Vietnam (Rodriguez and Preston 1999) with local and exotic (Tam Hoang breed from China) hens, in a semi-scavenging system with free night-time access to either fresh duckweed or broken rice offered separately, or mixed together, indicated that the local chickens ate much more fresh duckweed than the exotics, which preferred the mixture. These indications of ecotype (breed)*nutrition interaction were strengthened by another report from Cambodia in which the same exotic breed laid significantly less eggs than local birds, especially when no supplement was given (Khieu Borin, personal communication). The egg production of the exotics increased markedly when rice bran was fed and more so when they received rice bran mixed with fresh duckweed, reaching almost the egg production rate of the local birds that showed only minor responses to supplementation (Figure 17).

Duck production

Duckweed is perhaps so named because ducks were observed to consume it in the wild. The more omnivorous duck, compared with the chicken, appears to utilise it highly effectively under field conditions. On sewage farms in the New England territory of Australia, wild ducks so vigorously consumed the duckweed that they initially prevented the high growth rates needed to lower nutrients in the water to desired levels. In Vietnam duckweed produced on nutrients from animal and human waste is given fresh with cassava waste (Photos 10-13) to ducks. Duckweed provides both energy and protein and also a complement of essential minerals.

Duckweed (Lemna trisulca) has been shown to be able to replace 50% of the fish meal in a conventional diet for ducklings (Hamid et al., 1993) but its use as a major feed has not been considered. In Vietnam there are 30 million ducks raised annually These ducks traditionally scavenge their food supplies from the rice fields. They obtain a considerable amount of spilt grain, especially just after harvest and also consume insects, crustaceans and slugs and snails. Similar systems are well developed in Indonesia where ducks are induced to "graze" various land areas being led in groups after being trained at birth to follow some common object.

Ducks are preferred over poultry in these countries because they are more resistant to disease. Changing management conditions, particularly introduction of short rotation rice varieties and heavy use of pesticides and herbicides, are reducing the opportunities for ducks to scavenge in rice fields. Full- or semi-confinement systems are thus becoming more widespread. Typically the ducks in these systems are allowed limited scavenging, often on canals or ponds, and are supplemented with paddy rice and broken rice of low market value. Duckweed appears to have a potential role in these more intensive systems. In one study, fresh duckweed replaced roasted soya beans in amounts ranging from 0 to 100% of the supplementary protein in a basal diet of broken rice (Table 11).

Table 11: Mean intake of fresh foods by ducks (Men et al., 1995).

	Treatments
Intake (g/d)	0	30	45	60	100
Broken rice	82	78	83	82	92
Roasted soya beans	27	19	15	12	0
Duckweed (fresh)	0	496	499	505	566
Mineral premix	0.25	-	-	-	-
Total DM intake (g/d)	95	108	108	105	107
Total CP intake (g/d)	17	23	22	21	18
LWt gain (g/d)	26	29	28	27	28
FCR (g.DM/g gain)	3.7	4.2	4.2	4.1	4.2

It appears that fresh duckweed can totally replace soya beans as a protein source for a duck fattening system based on broken rice. Growth rates were not affected by level of inclusion of duckweed and there was only a slight deterioration in feed conversion at the highest level. The truly important issue here is that the research confirms the potential of duckweed as a high quality mineral source with protein equivalent to soya bean protein but which can be produced locally on the small farm.

Expensive mineral premixes were unnecessary when duckweed was added to the diet indicating a major practical and economic role of duckweed to provide the array of minerals for this level of production. Analyses of duckweed (Table 12) show that it is a well-balanced source of calcium and phosphorus with an appropriate array of trace elements. The importance of ensuring that the duckweed is of high nutritive value was made clear in work reported by Becerra et al (1995). In a basal diet of sugar cane juice, replacement of boiled soya beans with fresh duckweed reduced duck growth rate (Table 13). However, the duckweed contained only 26% protein in the dry matter. The water intake may also have been a constraint since Nguyen Duc Anh and Preston (1998) reported linear increases in growth rate of ducklings when increasing amounts of fresh duckweed were added to a basal diet of raw sugar and cassava root meal.

Table 12: Values for the composition of duckweed, broken rice and soya beans used in the studies of Men et al (1995,1996)

	Duckweed	Broken rice	Roasted soya bean
Dry matter (%)	4.7	86.8	87.0
Composition (% DM)
Crude protein	38.6	9.5	44.0
Ether extract	9.8	1.4	21.1
NFE	8.6	8.0	16.1
Fibre	18.7	2.0	9.8
Ash	19.0	1.1	5.6
Ca	0.7	0	0.2
P	0.6	0.2	0.9
K	4.3	-	-
Na	0.1	-	-
Fe	0.3	-	-
Mn (mg/kg)	1723	-	-
Zn (mg/kg)	75	-	-
Cu (mg/kg)	20	-	-
Carotene (mg/kg)	1025	-	-

Table 13: Effects of replacing boiled soya beans with duckweed in a diet of reconstituted sugar cane juice on intake and production of ducks (initial weight was 920g) (Becerra et al., 1995)

Replacement of soya bean (%)	0	15	25	35	45
Liveweight gain (g/d)	29.5	25.5	24.4	21.6	20.8
Feed intake (g DM/d)	147	154	149	145	138
CP (g/d)	25.7	27.3	24.8	22.2	19.8
Calculated H₂O from feed (g/d)	390	790	790	792	780

PIG PRODUCTION

Undoubtedly farmers have used aquatic plants to feed pigs in many countries but only a small amount of controlled research has been reported. Haustein et al. (1992a) fed pigs on a conventional grain-based diet and replaced part of the protein requirements with a low quality duckweed (23% CP with 7.5% fibre) harvested from a natural lake. In this instance the pigs were relatively young and as the duckweed meal increased in the diet, liveweight gain was significantly reduced. The results of this study are shown in Table 14.

Table 14: The effect of replacing "conventional" protein sources with duckweed meal (23% CP) in a concentrate based diet for pigs

	Level of duckweed in diet (%)
	0	5	10
Initial weight (kg)	6.9	6.8	6.8
Final weight (kg)	23.8	19.5	17.2
Live weight gain (g/d)	423	320	260

A number of issues may be raised with this study, the main one being that the duckweed meal was relatively low in protein (23% in dry matter) and the poor responses mirror those obtained by Becerra et al (1995) when low-protein duckweed was fed to ducks. Other problems may also have occurred, for instance, a high level of oxalate in the duckweed or the accumulation of some toxin or heavy metal is always a possibility. The disappointing result reported here should not detract from further studies.

In unpublished work from Instituto de Investigaciones Porcinas (Havana, Cuba) results from partial replacement (10%) of soya bean meal with duckweed meal (produced on the effluent from pig houses) in a basal diet of "B" molasses, demonstrated that small proportions of duckweed in the diet were used efficiently for growing pigs (Table 15).

Table 15: Effects of replacing soya bean meal (SBM) with duckweed on the growth of pigs over 3 months on a basal diet of molasses (Figueroa Vilda, unpunblished)

	Diet	Supplement	Liveweight gain (g/d)
Treatment 1	Molasses B	30% SBM	635
Treatment 2	Molasses B	SBM+duckweed*	630
Treatment 3	Molasses B	24% SBM	564

* 10% of the protein from dried duckweed and 24% from soya bean meal

More recently Rodriguez and Preston (1996b) fed three groups of genetically different pigs on unconventional diets containing duckweed. The experimental groups were local Mong Cai (MC) pigs, Large White (LW) or Mong Cai crossed with Large White (MCxLW). These were given sugar cane juice as the major energy source together with duckweed as a protein/mineral supplement. The pigs that had been highly selected for their abilities to utilise high grain/high quality protein diets (the 'so called' high-genetic potential Large White) were reluctant to eat the duckweed, whereas the native pigs and their crosses utilised duckweed efficiently, consuming significant quantities in addition to the free choice sugar cane juice. There was a linear increase in N retained, as the intake of dry matter from duckweed was increased as a proportion of the total diet. When duckweed represented about 50% of the total dry matter intake, 50% of the N from the duckweed was stored in tissues. Dry matter digestibility of the duckweed was predicted (from the regression of DM digestibility on percent duckweed in the diet) to be of the order of 66% (Figure 18). Rodriguez and Preston (1996b) suggested that the protein of duckweed is readily utilised and may have a higher biological value than meals such as those prepared from cassava leaves .

Figure 18: Relationship between percent of diet DM consumed as duckweed and apparent DM digestibility of Mong Cai (MC) or Mong Cai/Large White crossed (MCxLW) pigs.

Further evidence for the value of fresh duckweed in practical pig diets can be seen in the work of Du Thanh Hang (1998) done in villages in Central Vietnam. On five farms the pigs were fed the conventional diet based on brewer's grains, rice bran and cassava root meal or the same diets plus 2 kg/pig/day of fresh duckweed from ponds manured with dung from cattle. There were significant increases in growth rate due to supplementation with duckweed (Figure 19).

The data discussed above for pigs and poultry indicate that there may be some large differences in the composition or availability of nutrients from low protein duckweed (i.e. plants which are slow-growing) compared with fast-growing duckweed of high protein content. In particular, there are indications that the protein is of lower nutritive value (availability or essential amino acids or composition) when the duckweed is grown on water of low nutrient content.

RUMINANTS

Unlike monogastric animals, for which feed analysis is indicative of nutrient availability to the animal, ruminants through their fermentative digestive system modify virtually all the protein and carbohydrate in the feed they consume. The nutrients become available as volatile fatty acids (which are the major energy source), and amino acids (produced by enzymatic digestion of microbial cells that have grown and been washed from the rumen). Forage proteins are, in general, degraded to ammonia in the rumen and the animal depends on microbial protein for its essential amino acid supply. The efficiency of production is primarily dependent on the establishment of an efficient microbial ecosystem in the rumen. The potential use of duckweed in ruminant diets is twofold:

· as a mineral source to correct deficiencies in the diet for both rumen microbes and the animal

· as an ammonia source for the rumen microbes.

These two roles are largely confined to the rumen since an efficient microbial digestive system is dependent on a full complement of essential minerals and a high level of ammonia in the fluid. Deficiency of minerals and / or ammonia (which may be produced from supplemental non-protein-nitrogen sources or by the degradation of dietary protein) results in a lowered microbial growth in the rumen with inefficient growth of the microbial milieu. The consequences of low microbial growth is a reduced protein relative to energy in the nutrients absorbed (see Preston & Leng 1986 for review) and often lowered digestibility of forage and reduced feed intake.

Ruminants under feeding systems found in most areas of the world are often deficient in an array of nutrients required by the microbial fermentative digestive system. This is the case, particularly when consuming mature dry forages or crop residues (straws / stubbles) and at times agro-industrial by-products (e.g. sugar cane tops, molasses and fruit residues). Duckweed with its high mineral and protein content can provide an array of nutrients for the rumen microbes to function efficiently on such diets. In this way a small quantity of duckweed could replace the use of multi-nutritional supplements such as molasses-urea block licks (see Leng 1984).

Duckweed also has some potential as a dietary protein source that may be modified or may actually provide bypass protein that is required by productive animals to meet their extra requirements for essential amino acids (see Preston & Leng, 1986).

There are some preliminary research results where duckweed has been fed as a supplement to ruminants, but there is a need for a major research effort in this area to develop a clear definition of the strategic importance of duckweed as a N source, as a bypass protein source and as a source of essential minerals including S, P, Na, K, Mg and trace minerals.

A diet of duckweed plus maize silage diet (1:2) produced higher growth rates in Holstein heifers than one based on maize silage, concentrates and grass (Rusoff et al., 1978, 1980). More recently, Huque et al. (1996) have examined duckweed as a source of N and minerals for ruminant animals in Bangladesh, using nylon bag incubation techniques to study the breakdown of duckweed in the rumen of cattle. The duckweed contained about 30% crude protein in the dry matter. The results showed that, in cattle fed forage and concentrate, the potential degradation of duckweed dry matter in the rumen was very high; 85% (Spirodela), 72% (Lemna) and 93% (Wolfia). The protein of duckweed was highly soluble in rumen fluid: 24% (Spirodela), 42% (Lemna) and 18% (Wolfia). Overall, 80, 87 and 94%, respectively, of the protein was apparently degraded in the rumen. At high feed intakes there was apparently some potential for a small amount of the protein from duckweed to escape degradation in the rumen and provide essential amino acids directly to the animal.

It seems probable that dried duckweed will provide a readily fermentable protein source together with a rich mineral level needed for creating an efficient rumen for animals fed low protein forages such as straw. The extent that duckweed can correct mineral deficiencies in diets for ruminants will depend on the composition of the duckweed, which in turn depends on the level of minerals in the water where the duckweed is growing.

The major role of duckweed in ruminant diets is likely to be as a source of minerals and ammonia-N for the rumen. Future research should examine its strategic use to stimulate ruminant production on crop residues. In developing countries, ruminants often subsist on by-products of agro-industries and crop residues that are often (mostly) low in minerals and in fermentable N. In many situations duckweed would be a valuable resource to ensure that ruminants utilise these feeds effectively by providing a soluble N source (e.g. ammonia) needed by the cellulolytic organisms for protein synthesis and also a source of P and S which are essential for microbial growth and therefore for the animal (Preston & Leng, 1986).

From the research of Huque et al. (1996) it appears that duckweed would require treatment to protect its protein in order to produce a meal that will deliver protein to the intestines and produce a high protein to energy ratio in the nutrients absorbed that will further advance ruminant productivity from crop residues. The potential responses of cattle to both fermentable N (i.e. N sources that give rise to ammonia in the rumen such as urea or leaf proteins) and to bypass protein have been discussed in many publications. Research to develop methods to protect the protein on duckweed could lead to it being an alternative to multi-nutrient blocks for supplementing low-nitrogen diets fed to ruminants (Sansoucy, 1995)

Smith and Leng (1993) incubated duckweed meal in rumen fluid from sheep and found that it was rapidly fermented with the production of ammonia. Unfortunately, treatment by heat, formaldehyde or xylose - three methods that have been successful in increasing the bypass protein content of soya bean meal - had no effect on the rate of release of ammonia. However, these chemicals were only sprayed on the meal and it is probable that some heat is necessary to bring about protection of the protein. Duckweed protein, like terrestrial plant leaf protein, is not easily protected from rumen degradation by any presently known methodology.

FISH

Most intensive fish farms cultivate fish that have very high value on national or world markets. In these intensive systems the fish require feed with extremely high protein levels and a well-balanced array of essential amino acids. This is quite often provided by fish meals. These farming systems have achieved a high level of production but with high cost inputs (including feed, fresh water, and the prevention of pollution), which put them out of reach of the small-holder farmer. Duckweeds are not easily accommodated into such high technology systems, even though a dry meal with excess of 45% crude protein could theoretically be produced. However, it may be possible to blend duckweed meals with fish meal as a major protein source for intensive fish farming.

A better approach is to develop systems in which the duckweed is fed in the fresh form. Herbivorous fish are cultivated in many parts of the world. They include many varieties of carp and tilapia. These fish are often regarded as inferior in taste but they are of great benefit in the diets of poor people who are often financially restricted to largely vegetarian diets, which at times may be deficient in essential amino acids. Fresh duckweed is well suited to intensive production of herbivorous fish (Gaiger et al., 1984). It has been shown that duckweed is converted efficiently into liveweight by carp and tilapia (Hepher & Pruginin, 1979; Robinette et al., 1980; van Dyke & Sutton, 1977, Hassan & Edwards 1992, Skillicorn et al., 1993).

The results of the duckweed research programme in Bangladesh (see later) has focussed world attention on duckweed both as a feed for freshwater fish and as an element in water purification (Skillicorn et al., 1993). The book on this subject is mandatory reading for anyone becoming interested in this area. The PRISM group initiated the pilot project in Bangladesh to develop farming systems for duckweed and to test its value as a fish feed in poly-carp production. The outcomes of this project point the way for the efficient use of duckweed in many situations. It has important lessons for the development of small-holder farmer systems where integration of crop and animal production can make efficient use of the capacity of duckweed to scavenge and retain major mineral nutrients within the system.

Carp production

Introduction

Carp species are by far the most commonly cultivated freshwater fish in Asia. They grow under diverse conditions and tolerate water of poor quality, such as is found in stagnant ponds. Their greatest attribute is the ease with which they can be managed in ponds and the huge production potential under good environmental and nutritional conditions.

Different carp species tend to occupy different ecological niches and therefore a number of species that are highly selective in their dietary preferences can be placed in the same pond and will occupy different feeding zones. The development of poly-carp culture depends on using all the feed biomass in a pond system by having "top", "middle" and "bottom" feeding fish. These include:

herbivorous species capable of feeding on surface plants or plants accessible on the sides of the pond, or fed to them freshly harvested from another site as would be the case with duckweed;
middle feeders which feed on either zooplankton and / or phytoplankton that grow on the detritus produced by the top feeders and
a bottom feeding species which uses the faecal materials produced by the middle and upper feeding fish.

Carp poly-culture depends on most of the fish being produced on the zooplankton and phytoplankton (up to 85%). Providing extra feed for the surface feeders may increase production. This was a major reason for establishing the PRISM projects at Mirzapur as the surface feeders were restricted to very low stocking as the only available plant biomass was from the pond edges.

The poly-carp system depends on reducing biochemical oxygen demand (BOD) so as to maintain high oxygen levels in the water. Aeration and fertilisation are critical factors in such a system. BOD is created by:

· high densities of phytoplankton that respire at night

· the high oxygen demand from the heavy stocking rate of fish

· aerobic microbial degradation of organic matter.

In a traditional manure-fed pond for poly-culture of carp the main consideration is at what rate should the manure be added? Stocking rate then is determined by the resultant demand for oxygen.

The special features of using duckweed for carp production are that the system can be set up using a single species of surface feeding carp or it can be used in poly-culture to increase the total stocking rates. By providing fresh duckweed daily, which does not decompose and can be fed ad lib, the top feeding carp densities can be increased together with an increase in bottom feeders thus increasing stocking density overall.

As the biochemical oxygen demand is less when there is a reduction in aerobic degradation of plant materials (i.e. replacing dead plant biomass with live duckweed), the fish levels can increase to an extent that their respiration approaches the oxygen needs for the degradation of the organic matter entering the pond. The incremental production of the top feeders (Grass, Catla and Mirror carps) and bottom feeders (Mrigal carp) represents the potential extra production from the duckweed system (Skillicorn et al., 1993).

Duckweed-fed carp poly-culture

The reader is referred to Skillicorn et al. (1993) for the most authoritative discussion of this subject. In this document I will discuss the main issues relating to the work of these authors. The major criticism of the work of Skillicorn et al. is that it is too sophisticated to be applied by a small-holder farmer. Resource-poor farmers need considerable economic support to set up fish farming. It becomes more complex where there is an attempt to integrate the farm in order for it to be sustainable. Under these conditions the use of artificial fertiliser, as in the Mirzapur project, limits the application by small-holder farmers. Fertilisers are often too expensive to use even for rice crop production. For example, much of the rice grown on small farms in Vietnam depends on recycling of nutrients via pig manure and fertilisers are either not used or are used sparingly. The financial crisis in Asia is likely to make fertilisers more expensive and it can be anticipated that there will be decreasing grain production in Asia over the next few years.

Despite these reservations about the approach, Skillicorn et al. (1993) have provided valuable data, which can provide guidelines for development of small-holder farming systems.

The Mirzapur carp stocking strategy and carp growth rates

Grass carp (Ctenopharyngodon idella) are the major users of duckweed but Catla (Catla catla) and Common carp (Cyprinus carpio) compete aggressively for duckweed. Only about 50% of the potentially digestible nutrients in duckweed are used by the fish and so the faeces from carp feeding on duckweed are relatively high in organic materials useable directly, or indirectly through microbial action, by the bottom feeders . As a result the bottom feeding species can be increased to 30% of the total population (see Skillicorn et al. 1993).

The distribution of carp in the Mirzapur venture (1989-90) was:

· 45% top feeders (15% Catla, 20% Grass Carp, 10% Mirror Carp)

· 35% middle feeders (Rohu 15%, Silver Carp 20%)

· 20% bottom feeders (20% Mrigal Carp)

The initial population was about 23,000 carp fingerlings per hectare. Yields were difficult to estimate but it was claimed that around 10 tonnes per hectare were produced annually. The weight of fish captured per month is shown in Figure 20. Weights of the fish are in Figure 21. The feed conversion ratio efficiency was estimated to be between 10 and 12 kg fresh duckweed to 1 kg of weight gain. This implies that other sources of feed were available to the fish. A major cost was the daily fertiliser application to the duckweed ponds. The growth rates of the fish showed that it was the middle feeders (Silver and Rohu Carp) that had the fastest growth rates. Grass carp grew disappointingly slowly. The average fish weights after 13 months feeding of duckweed to the poly-culture are shown in Figure 21.

Figure 20: Average weight of fish by month in the Mirzapur trials

Table 16: Weight of different carp species in poly-culture fed with duckweed

Species of carp	Weight (kg) at		Calculated
	8 months	13 months	growth rate (g/d)
Catla	0.9	1.9	1.0
Grass carp	0.7	0.8	0
Mirror carp	1.5	0.9	-0.6
Rohu	0.6	1.5	0.9
Silver carp	1.4	2.7	1.3
Mrigal	0.5	0.4	-0.1

Figure 21: Average weight of fish species after 13 months in Mirzapur

The average growth rate of the carp in poly-culture (Figures 20 and 21 and Table 16) is calculated on the assumption that the fingerlings weighed only a few grammes when placed in the ponds and that this is insignificant in relation to the final weight.

Feeding duckweed

In the Mirzapur venture, fresh duckweed was the only supplementary "feed" provided to the fish. The rate of feeding was set at a level that would result in a slight excess after feeding activity had finished. With 30,000 fish per hectare there were no problems of low oxygen levels in the water. The duckweed was fed either in enclosed (surface) feeding stations or was simply tipped into the pond at the edges.

The data on fish growth as reported by Skillicorn et al. (1993) are unfortunately compromised by the "normal" problems encountered by research workers who set up "practical" demonstrations and then attempt to analyse the observed data. There were serious logistical problems, at least initially, in providing the feed consistently. The fish sampling techniques were based on frequent harvesting with removal of the largest and smallest fish at each harvest. The large ones were presumed to be reaching the weight at which their growth rate slows whereas the small ones were deemed to be "poor growers".

The studies showed that under the conditions of the trials Silver carp, Catla and Rohu grew significantly more rapidly than the others. However, problems of inadequate supply of duckweed compromised the growth of the herbivorous species. In other studies when they were fed duckweed ad libitum, they grew to 4kg in six months (20g/day) as compared with 1 kg in 13 months in the poly-culture. This suggests that the production of grass carp may be more appropriately promoted as a monoculture.

Some ideas on duckweed use for carp production at small-holder farm level

Carp poly-culture in ponds fertilised with manure is an efficient and established method of producing carp. However, it is somewhat specialised and not managed easily by small-scale farmers with limited resources of land and nutrient-enriched water. A further constraint is the capital outlay for the relatively deep ponds (2-3 m) that are required.

If grass carp are capable of high growth rates when supplemented with ad lib duckweed, as shown by the PRISM project, then it could be advantageous to move away from poly-culture to the culture of single species perhaps produced in sequence in the same pond. Thus the first stage would be to produce Grass carp, feeding them with duckweed. The Grass carp would then be removed from the pond and Rohu and Silver carp would be introduced to utilise the phytoplankton/zooplankton that build up and then the pond would again be fed with duckweed to extract the nutrients liberated from the detritus. The detritus feeders, not being very productive, would not be included in such an enterprise.

Thus, a potential strategy (Figure 22) would be to periodically reverse the roles of the duckweed pond and the fish pond. This would be especially effective where effluent low in organic matter is used as the medium for duckweed culture, particularly the liquid from a biodigester; and where water is scarce and therefore must be used efficiently.

Trial and error would be needed to work out a routine but a starting point could be to have ponds of equal size. At any one time, one pond could be producing duckweed and the other Grass carp. Three ponds could also be used. One (DW) of these would receive the effluent from a biodigester and the other two would produce fish: pond Pond GC would receive the duckweed from pond DW and pond PZ would be used for fish feeding on phytoplankton and zooplankton. The ponds could be rotated but maintaining the sequence of DW, GC, PZ, DW, GC …etc. In this way each pond is used, in turn, for each activity, so that always the duckweed can reuse the minerals released by the faeces of the fish. The grass carp in the GC pond would be kept there until the oxygen concentration in the water had declined to levels where the fish ceased to grow. Then plankton-consuming carp (PZ) could be introduced. Duckweed is grown to lower the mineral content of the pond water, after which the duckweed pond (DW) then reverts to a fish pond (GC) for Grass carp. Such a system would allow the use of static or slow-moving water with a relatively slow rate of turnover.

The only major water loss from the system would be by evaporation, if seepage and use of water for other purposes are minimised. The operation of the three pond system over 18 months would be as indicated in the table 17.

Table 17: Monoculture of carp using duckweed in an integrated system


Months	Pond 1	Pond 2	Pond 3
1-6	Duckweed	Grass carp	Silver carp
7-12	Grass carp	Silver carp	Duckweed
13-18	Silver carp	Duckweed	Grass carp

Figure 22: Outline of a scheme for production of carp in a three pond system. Duckweed from pond 1 feeds the top-feeding carp in pond 2 and plankton-feeding carp are placed in pond 3 after the grass carp have been harvested. The ponds can then be rotated in their use.

The success of such a project would depend on:

· relatively clean water emerging from the duckweed pond.

· the Grass carp having grown to a saleable size before the BOD becomes critical.

· the continuous growth of plankton on the nutrients in the faeces from the previous crop of Grass carp but before oxygen levels in water become limiting. Stocking rates would need to be adjusted to allocate sufficient feed for the Silver carp to reach market weight.

· low organic matter in the inlet to the duckweed pond.

Where water is scarce, either year round or seasonally, this system could have certain advantages. However, some factors would need researching such as:

the numbers of animals (large or small ruminants) needed to provide the manure for the biodigester
the degree of reduction of the BOD of the manure in passage through the biodigester.

Initially the findings in the Mirzapur venture regarding NPK fertiliser needs would give some guidelines. A lagoon producing 1000 kg fresh duckweed per ha per day received about 20 kg of urea, 4 kg of trisodium phosphate, 4 kg of muriate of potash and 9 kg of sea salt per hectare of the surface area. Of these nutrients probably ammonia (from the urea) is the nutrient that needs to be most carefully controlled as it can quickly limit duckweed production above 60 and below about 10 mg N/litre.

If a cow is consuming ammoniated rice straw supplemented with minerals and some protein meal, the manure (faeces and urine) probably contains only slightly less than the N present in the ration. A cow would consume 2.5% of its body weight as dry matter. Thus a cow weighing 400 kg would consume 10 kg of feed dry matter and produce 5 kg of manure dry matter containing most of the N fed as ammonia in the straw.

If 4% urea had been used to treat the straw then in 10kg of straw there would be the equivalent of 200g of urea (half the ammonia will be lost between treatment and consumption) in the washings (say 80 litres) from the stall. So there is about 25 mgN/litre. This is a little lower than the level recorded in the laboratory for the optimal growth of duckweed. With a number of cows producing faecal and urinary N at this rate it would mean that 10 cows would produce 2 kg urea and therefore 100 cows could fertilise 1 ha or 10 cows for 0.1 ha. of duckweed lagoon.

Production of tilapia

Tilapia can use duckweed efficiently when fed at an appropriate rate. In studies in Thailand (Hassan and Edwards 1992), tilapia were grown in static water in concrete tanks and fed two species of duckweed (Lemna perpusilla and Spirodela polyrhiza) at various levels of duckweed dry matter per kg of fish (fresh weight) per day. The duckweed was relatively low in protein (approx. 24% CP). The Spirodela was poorly consumed whereas Lemna was rapidly ingested by the fish (Table 18).

Table 18: The effects of feeding tilapia increasing levels of Lemna. The tilapia initially weighed approximately 41 g. (Hassan and Edwards 1992).

Feeding rate	Survival	Mean live	Conversion of Lemna
of Lemna	rate of fish	weight gain	DM to fish live weight
(g DM/kg fish)	(%)	(g/d)	(g/g)
10	97	0.2	1.9
20	100	0.4	1.9
30	100	1.0	1.6
40	60	1.0	2.3
50	27	0.7	3.3
60	17	0.8	3.3

The high death rates when more than 30g Lemna dry matter per kg fish were fed daily presumably was due to the eutrophication of the water. Over-consumption of duckweed, however, cannot be ruled out as a cause of the high death rate.

Where duckweed is fed to tilapia in open waters or lagoons it is fed fresh and does not appear to effect the biochemical oxygen demand directly as the rate of feeding is controlled by the farmer so as to maintain only small amounts of excess duckweed on a daily basis (personal observations).

The work of Hassan and Edwards (1992) indicates the levels of consumption of duckweed by Tilapia and its potential role as a feed that can be added to relatively unpolluted water. The growing importance of farmed Tilapia suggests that a greater research effort is needed to develop inexpensive systems in which Tilapia densities can be increased by addition of extra feed without reduction in size of fish. Tilapia could replace the top feeding carp in the systems proposed in Figure 22, retaining the rotation of phytoplankton-consuming carp and the production of duckweed.

Indirect use of duckweed to produce feed for fish

Ogburn and Ogburn (1994) developed an oxidation treatment of sugar-mill waste water using duckweed that appeared to be highly successful. The work was carried out in Negros Oriental, in the Philippines. The mean ammonia concentration in the influent, relative to effluent water, from the treatment plant over a six month period was reduced before release to the ocean from 0.87 to 0.31mg ammonia/litre. Comparable data for orthophosphate were from 0.93 to 0.5 mg P₂0₅/litre and for biochemical oxygen demand from 611 to 143 mg BOD/litre. The reduction in pollutants in the waste water over the three years since the inception of the treatment plant apparently prevented the mass death rates of fish that occurred annually in the bay that received the water. Duckweed production was 8.8g/m²/day. It was observed that "milk-fish" (Chanos chanos) did not consume duckweed. The approach that was taken was to harvest the duckweed and apply it to the base of the pond prior to introduction of water. The dead duckweed fertilised the production of lablab, which the fish consumed. "Lablab" is Filipino for the biological complex of blue green algae, diatoms, bacteria and animals that form on the bottom or float to the surface of ponds. Fertilising fish ponds in this way, as compared to inorganic fertiliser or cow manure, enhanced fish production as shown in Table 19 (Ogburn & Ogburn 1994).

Table 19: The effects of different fertiliser application on the production of milk fish (Chanos chanos).

Lagoon system for	Milk fish harvested
providing feed	(kg/ha/90d)
1. Inorganically fertilised	320
2. Fed with cow manure	545
3. Fed duckweed	820

DUCKWEED AS A SOURCE OF NUTRIENTS FOR HUMANS

Duckweed has been used as a food by poor people in the past. The major benefit from such an addition to a diet is likely to have been as a supplement rich in phosphorus and / or vitamin A. However, undoubtedly there is a role for duckweed as a source of essential amino acids. Duckweed can be added to a salad and apparently is quite tasty.

Where vegetable proteins are scarce in some regions of the world and particularly during a prolonged dry season, or in normally arid areas, there may be considerable scope to improve the nutritional status of mal-nourished children through the consumption of duckweed, directly or after extraction of the protein. Many aquatic plants may be used for such purposes with some additional purification to remove any toxic materials and also reduce the level of poly-phenols. Dewanji (1993) demonstrated this effectively with several aquatic weeds in India. The weeds were subjected to pulping and filtration to extract mainly protein which was precipitated by steam injection. The chemical composition of such a protein extract is shown in Table 20 and the amino acid composition in Table 21.

Table 20: Composition of protein extracts from three common aquatic weeds (modified from Dewanji & Matai, 1991)

	*Azolla*	*Lemna*	*Pistia*
N (%)	6.3	6.1	8.2
Crude fat (%)	9.9	11.4	14.4
Crude fibre (%)	2.8	2.7	1.5
Ash (%)	4.1	6.0.	5.8
b-carotene (mg/g)	632	627	654
Poly-phenols (%)	1.7	2.1	1.3
In vitro digestibility (%)	78	78	81

As a source of essential amino acids the protein of water plants has a comparable amino acid composition to that of most leaf proteins. The protein extract would provide quite considerable benefits to communities constrained to vegetarian diets because of their economic situation. This would particularly apply to those without a source of milk and where there is a long period of dependency on dried foodstuffs deficient in vitamin A or in phosphorus as occurs in many of the arid regions of the world. On the other hand, with the increasing demand for vegetable proteins in the industrialised world duckweeds could make a fine addition to most mixed salads and could be regarded as a commercial crop, provided water of good quality was used to grow the plants.

Table 21: Amino acid patterns (g/100g protein) in leaf protein extracted from three aquatic weeds compared with leaf protein from alfalfa (see Dewanji 1993 for details)

Amino acid	*Azolla*	*Lemna*	*Pistia*	Alfalfa
lysine	6.1	5.9	7.0	6.7
histidine	2.3	2.7	2.9	2.5
argenine	6.2	6.0	6.3	6.5
aspartate	1.3	10.6	9.6	10.2
threonine	5.0	5.1	4.8	5.2
serine	5.3	5.4	4.8	4.3
glutamic acid	13.8	13.6	13.4	11.1
proline	4.7	4.5	5.0	4.8
glycine	5.8	5.6	5.7	5.3
alanine	7.0	7.1	6.3	6.0
valine	6.8	6.4	6.7	6.8
methionine	1.2	1.4	1.1	2.3
isoleucine	6.0	5.7	5.9	5.3
leucine	9.4	9.6	9.6	8.9
tyrosine	4.2	4.2	4.6	4.4
phenylalanine	5.9	6.0	6.0	5.7

Safety considerations when duckweed enters the human food chain

Whilst there appears to be considerable scope to use duckweed as components of diets for animals and to a small extent for humans, it is necessary to be cautious in recommending wide scale application. The nature of duckweed as a scavenger of minerals poses potential danger where heavy metal contamination of water occurs.

This is of increased concern where waste materials from nuclear reactors have leaked into the environment. These may also be concentrated in duckweed. Heavy metals can enter the food chain at a number of points and it needs to be stressed strongly that there is a need to monitor the levels of heavy metals in duckweed that may be produced in large-scale operations for any food/feed purposes.

Chapter 8: References

Abdulayef D.A. 1969. The use of common duckweed as green feed for chickens. Uzbekskii Biologiheskii Zournal (USSR). 13:42-43.

Ashbey, E. & Wangermann, E. 1949. Senescence and rejuvenation in Lemna minor, Nature 164:187

Becerra, M. 1994. Evaluation of feeding systems for growing ducks based on aquatic plants and sugar cane juice. M.Sc. Thesis, Swedish University of Agricultural Sciences Uppsala 1994.

Becerra, M. Preston, T.R. & Ogle, B 1995. Effect of replacing whole boiled soyabeans with Lemna sp in diets of growing ducks. M. Sc. Thesis, Swedish University of Agricultural Sciences.

Bellini, R., Veronesi, R & Rizzoli, M 1994. Efficacy of various fish species (Carassius auratus Cyrprinuscarpio, Gambusia affinis) in the control of rice field mosquitoes in Northern Italy. Bulletin of the Society for Vector Ecology 19(2):87-99

Crombie, L & Heavers, A.D. 1994. Synthesis of the algicidal alelochemical from Lemna trisulca (Duckweed) Journal Chemical Society, Perkin Transactions 10:2683-2687

Culley, D.D. & Epps, E.A. 1973. Use of duckweed for waste treatment and animal feed. Journal of Water Pollution Control Federation. 45(2): 337-347.

Datko, A.H. & Mudd, S.H. 1984. Sulphate uptake and its regulation in Lemna paucicostata (Hegelin 6746) Plant Physiology 75:466-473

Dewanji, A. & Matai S. 1991. Nutritive value of extracted leaf proteins from selected aquatic plants. Indian Journal of Agricultural Biochemistry 4:61-65

Dewanji, A. 1993. Amino acid composition of leaf proteins extracted from some aquatic weeds. J. of Agriculture and Food Chemistry 41:1232-1236

Dewanji, A & Matai, S. 1996. Nutritional evaluation of leaf protein extracted from three aquatic plants. Journal of Agricultural and Food Chemistry 44(8):2162-2166.

Devi, M. Thomas, D.A., Barber, J.T. & Fingerman, M. 1996. Accumulation and physiological and biochemical effects of cadmium in a simple aquatic food chain. Ecotoxiology and Environmental Safety 33:38-43

Du Thanh Hang 1998 Ensiled cassava leaves and duckweed as protein sources for fattening pigs on farms in Central Vietnam. Livestock Research for Rural Development (10) 3: http://cipav.org.co/lrrd/lrrd10/3/hang2

Duong, T.P. & Tiedje, J.M. 1985. Nitrogen fixation by naturally occurring duckweed - cyanobacterial associations. Canadian Journal Microbiology 31:327-330

Eid, M.A.A. Kandil, M.A.E., Moursy, E.B. & Sayed, G.E.M. 1992a. Bioassay of duckweed vegetation extracts. Insect-Science, and its application 13:741-748

Eid, M.A.A. Kandil, M.A.E., Moursy, E.B. & Sayed, G.E.M. 1992b. Effect of duckweed, Lemna minor vegetations on the mosquito Calex bibiens pipiens. Insect Science and its Application 13(3):357-361

Epstein, P.R. 1993. Algal blooms in the spread and persistence of cholera Biosystems 31:209-221

FAO/WHO 1991. Protein quality evaluation. Report of a joint FAO/WHO expert consultation. FAO Food and Nutrition Paper 51, FAO Geneva 1991

FAO 1989. Production year book FAO Rome

Fleay B.J. 1966. The decline of the age of oil. Pluto Press Australia.

Gaiger, I.G., Porath, D. & Granoth, G. 1984. Evaluation of duckweed (Lemna gibba) as feed for tilapia (Orechromis niloticus cross Orechromis aureus) in a recirculating unit. Aquaculture, 41, 235-244

Hamid, M.A., Chowdhury, S.D. Razzak, M.A. & Roy, C.R. 1993. Effects of feeding an aquatic weed Lemna trisulca as partial replacement of fish meal on the performance of growing ducklings. Journal of Science of Food and Agriculture 61(1):137-139

Hanczakowski, P., Symczyk, B. & Wawrzyñski, M. 1995. Composition and nutritive value of sewage-grown duckweed (Lemna minor) tarrats Animal Feed Science and Technology 52, 339-343

Hassan, M.S. & Edwards, P. 1992. Evaluation of duckweed (Lemna perpusilla and Spirodela polyrhiza) as feed for Nile Tilapia (Oreochromis niloticus) Aquaculture 104:315-326

Haustein, A.T., Gillman, R.H., Skillicorn, P.W., Vergara, V., Guevara, V., & Gastanaduy, A. 1980. Duckweed, A useful strategy for feeding chickens: Performance of layers fed with sewage-grown Lemnacea species. Poultry Science. 69: 1835-1844.

Haustein, A.T., Gonzalez, Gillman, R.H., Campos, K., Caldas, M., Armas, F., Madrid, C., & Castro, M., 1992a. Uso de Lemnaceae en la alimentacion del ganado porcino. Asociación Benéfica PRISMA - Facultad de Medicina Veterinaria, Universidad Nacional Mayor de San Marcos

Haustein, A.T., Gillman, R.H., Skillicorn, P.W.,Guevara, V., Dias, F., Vergara, V., Gastanaduy, A. & Gillman, J.B. 1992b. Compensatory growth in broiler chicks fed on Lemna gibba. British Journal of Nutrition. 68(2):329-335.

Haustein, A.T., Gillman, R.H., Skillicorn, P.W., Hannan, H., Dias, F., Guevara, V., Vergara, V., Gastanaduy, A. & Gillman, J.B. 1994. Performance of broiler chickens fed diets containing duckweed (Lemna gibba). Journal of Agricultural Science 122(2):288-289

Hepher, B., & Pruginin,Y. 1979. Guide to fish culture in Israel. 4. Fertilization, manuring and feeding. Foreign Training Dept., Israel. 61pp.

Hogan, K.B. 1993. (ed) Opportunities to reduce anthropogenic methane emissions in the United States (Report to Congress) US Environmental Protection Agency, Office of Air Radiation (EPA 430-R-93-012)

Huque, K.S., Chowdhury, S.A., & Kibra, S.S., 1996. Study on the potentiality of duckweeds as a feed for cattle Asian-Australasian Journal of Animal Sciences 9, 133-137

Landolt, E., & Kandeler, R. 1987. Biosystematic investigations on the family of duckweeds (Lemnaceae) published in the Veröffentlichungen des Geobotanischen Instituties der ETH, Stiftung Rübel, Zürich

Lautner, V & Muller, Z., 1954. Feeding value of some water plants I. (In Czech.). Sb. Cesk. Akad. Zemedl. Ved 27A:333-354.

Le Ha Chau 1998 Biodigester effluent versus manure, from pigs or cattle, as fertilizer for duckweed (Lemna spp.). Livestock Research for Rural Development (10) 3: http://www.cipav.org.co/lrrd/lrrd10/3/chau2

Leng, R.A., Stambolie, J.H., & Bell, R.E. 1994.Duckweed a potential high protein feed resource for domestic animals and fish In: Improving animal production systems based on local feed resources. 7th AAAP Animal Science Congress 100-117

Leng, R.A. 1984. The potential of solidified molasses-based blocks for the correction of multi-nutrient deficiencies in buffaloes and other ruminants fed low quality agro-industrial byproducts. In: "The use of nuclear techniques to improve domestic buffalo production in Asia" pp 135-150 Vieena: International Atomic Energy Agency.

Marten, G.G., Suarez M.F., & Astaeza, R. 1996. An ecological survey of Anopheles albimanus larval habitats in Colombia. Journal of Vector Ecology 21(2):122-131

Mastika, I.M. and Cumming, R.B. 1985. Effect of nutritional and environmental varieties on choice feeding broilers. Recent Advances Animal Nutrition in Australia in 1985 UNE publishing Unit, Armidale, NSW.

Maywald, A., Zeschmar-Lahl, B., & Lahl, U. 1988. Water fit to drink in Earth Report Beazley M. (edit), Mitchell Beazley Publish.

Mbagwu, I.G. & Adeniji, H.A. 1988. The nutritional content of duckweed (Lemna paucicostata Hegelm). in the Kainji Lake area, Nigeria, Aquatic Botany 29, 357-366

Men, X.B. Ogle, B., & Preston, T.R., 1995. Use of duckweed (Lemna spp) as replacement for soyabean meal in a basal diet of broken rice for fattening ducks. In Livestock Research for Rural Development 7:5-8

Men, X.B. Ogle, B., and Preston, T.R., 1996. Duckweed (Lemna spp) as replacement for roasted soya beans in diets of broken rice for fattening ducks on a small scale from the Mekong delta. Livestock Research for Rural Development 8, 14-19

Mesmar, M.N. & Abussaud, M 1991. The antibiotic activity of some aquatic plants and algal extracts from Jordon. Qatar University Science Bulletin 11:155-160

Mestayer, C., Culley, D.D., Standifer, L.C., and Koonce, K.L., 1984. Solar energy conversion efficiency and growth aspects of the duckweed, Spirodela punctata Aquatic Botany. 19:157-70

Muzaffarov, A.M. 1968. The use of common duckweed for feeding domestic birds. Uzebekskii Biologicheskii Zournal (USSR) 12(3):42.

National Academy of Sciences - National Research Council 1976. 'Making aquatic weeds useful: Some perspectives for developing countries' Washington, D.C. Waste water treatment using aquatic weeds, Chapter 13:115-174

Nguyen Duc Anh, Huu Tao and Preston, T. R. 1997. Effect of management practices and fertilization with biodigester effluent on biomass yield and composition of duckweed. Livestock Research for Rural Development (9) 1: http://www.cipav.org.co/lrrd/lrrd9/1/anh91

Nguyen Duc Anh and Preston, T. R. 1997. Evaluation of protein quality in duckweed (Lemna spp.) using a duckling growth assay. Livestock Research for Rural Development (9) 2: http://www.cippaav.org.co/lrrd/lrrd9/2/anh92

Nolan, J.V., Bell, R.E., Thomson, E., Bremner, D. & Ball, W. 1997. Duckweed (Spirodela punctata) as a protein and pigment source in diets for layers. Proceeding of Australian Poultry Science Symposium 9, 166-170

Noraho, N & Gour, J.P. 1996. Cadmium absorption and intracellular uptake by two macrophytes, Azolla pinnata and Spirodela polyrhiza. Archiv. fuer Hydrobiologic 136(1):135-144

Ogburn, D.M. & Ogburn, N.J. 1994. Use of duckweed (Lemna spp) grown in sugar-mill effluent for milk fish, (Chanos chanos Forskal) production. Aquatic and Fisheries Management 25(5):497-503

O'Neil, P.J., Nolan, J.V. & Thomson, E. 1996. Duckweed as an alternative to soyabean meal in diets for high producing layers. Proceedings Australian Poultry Science Symposium 8:123-127

Oron, G., Porath, D. & Jansen, H., 1987. 'Performance of the Duckweed species Lemna gibba on municipal wastewater for effluent renovation and protein production' Biotechnology and Bioengineering XXIA: 258-269 (John Wiley & Sons Inc.)

Porath, D., Hepher, B. & Koton, A. 1979. Duckweed as an aquatic crop: Evaluation of clones for aquaculture. Aquatic Botany 7:273-278.

Preston, T.R., & Murgueitio, E. 1992. Strategy for sustainable livestock production in the tropics SAREC. Publishers Consultorias para el Desarrollo Rural Integrado en el Tropico (CONDRIT) Ltda, Cali, Colombia.

Preston, T.R. 1990. Future strategies for livestock production in tropical third world countries. Ambio 19:390-393

Preston T.R. 1995. Sugar cane for feed and fuel: recent developments: Review Monrdial Zootechnica 82:84-89

Preston, T.R. & Leng, R.A. 1986. Matching Livestock Systems to Available Resources in the Tropics and Subtropics. Penambul Books, Armidale Australia.

Reddy, K.R. & DeBusk, W.F. 1985. Growth characteristics of aquatic macrophytes cultured in nutrient enriched water: II Azolla, Duckweed and Salvinia. Economic Botany 39(2):200-208

Robinette, H.R., Brunson, M.W. & Day, E.J. 1980. Use of duckweed in diets of channel catfish. Proc. 13th Ann. Conf. SE Assoc. Fish Wildlife Age 108-114

Rodriguez, L. & Preston, T.R. 1996a. Use of effluent from low cost plastic Biodigesters as fertilizer for duckweed ponds. Livestock Research for Rural Development 8:60-66

Rodriguez, L. & Preston, T.R. 1996b. Comparative parameters of digestion and N-metabolism in Mong Cai/Large white cross piglets having free access to sugar cane juice and duckweed. Livestock Research for Rural Development 8:72-81

Rodriguez L. & Preston, T. R. 1999. Observations on scavenging local (indigenous) and Tam Hoang (exotic) chickens given free access (when confined at night) to duckweed (Lemnaceae) offered alone or mixed with rice bran. Livestock Research for Rural Development (9) 2: http://www.cipav.org.co/lrrd/lrrd11/1/lyl111.htm

Rusoff, L.L., Zeringue, S.P., Achacoso, A.S. & Culley, D.D. 1978. Feeding value of duckweed (an aquatic plant, Family Lemnaceae) for ruminants. The American Dairy Science Association Michigan State University, East Lansing, Mich. July 9-13.

Rusoff, L.L. ,Blackeney, E.W. & Culley, D.D. 1980. Duckweed (Lemnaceae family): a potential source of protein and amino acids. Journal of Agriculture and Food Chemistry. 28: 848-850

Sansoucy, R. 1995. New developments in the manufacture and utilization of multi-nutritional blocks. Review Mondial Zootechnica 72:82-83

Sell, K.L. Burton, J.W. & Leng, R.A. 1993. Unpublished reported by

Sell, K.L. 1993. An investigation into the performance of duckweed (Lemnacae) as a biological method of salt removal from mine water. Thesis B. Nat.Res. University of New England, Armidale 2351 Australia.

Skillicorn, P., Spira, W. & Journey, W., 1993. Duckweed Aquaculture - A New Aquatic Farming System for Developing Countries. The World Bank. 76pp. Washington DC.

Smith, P., & Leng, R.A. 1993. Rural Science Honours thesis "Bypassing Duckweed Protein" U.N.E. - Armidale NSW, Australia.

Stambolie, J.H. & Leng, R.A. 1994. Unpublished observation U.N.E. Armidale NSW, Australia

Sutton, D.L. & Ornes, W.H. 1975..Phosphorous removal from static sewage effluent using Duckweed. Journal of Environmental Quality 4(3):367-370.

Sutton, D.L. & Ornes, W.H. 1977. Growth of Spirodela polyrhiza in static sewage effluent. Aquatic Botany 3:231-237

Truax, R. Culley, D.D., Griffith, M., Johnson, W. & Wood, J. 1972. Duckweed for chick feed? Louisiana Agriculture 1 8-9

Van Hove, C. 1989. Azolla and its multiple uses with emphasis on Africa FAO.

Van-Dyke, J.M. & Sutton, D.L. 1977. Digestion of duckweed (Lemna spp) by the grass carp (Ctenopharyngodon ldella). Journal of Fish Biology 11:273-278.

Vatta, G., Rota, R., Boniardi, N. & Nano, G. 1995. Dynamic modelling of waste water treatment plants based on Lemna gibba. The Chemical Engineering Journal 57: 1337-1348

Wangermann, E., & Ashby, E. 1950. Morphogenesis in Lemna minor. Proceeding Linn. Soc. London 162:10-13

Wolff, P. & Landolt, E. 1994. Spread of Lemna turionifera (Lemnaceae), the red duckweed in Poland. Fragmenta Floristica et Geobotanica 39, 439-451

World Watch. 1997. -World Watch Institute 1997

[1] Based in Colombia, Maryland, USA, see chapter 6