A NEW AQUATIC FARMING SYSTEM FOR DEVELOPING COUNTRIES Paul Skillicorn, William Spira, and William Journey 4 n ~~~~~~~~ ' 4L~ _ A WORLD BANK PUBLICATION ~~~~~~~~~~~ -0 DUCKWEED AQUACULTURE A NEW AQUATIC FARMING SYSTEM FOR DEVELOPING COUNTRIES Paul Skillicorn, William Spira, and William Journey THE WORLD BANK WASHINGTON, D.C. Copyright (C 1993 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing March 1993 The flndings, interpretations. and conclusions expressed in this paper are entirely those of the authors and should not be attributed in any manner to the World Bank. to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. Any maps that accompany the text have been prepared solely for the convenience of readers: the designations and presentation of material in them do not imply the expression of any opinion whatsoever on the part of the World Bank, Its affi]iates. or its Board or member countries concerning the legal status of any country, territory, city or area or of the authorities thereof or concerning the delimitation of its boundaries or its national affiliation. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and. when the reproduction is for noncommercial purposes. without asking for a fee. Permission to photocopy portions for classroom use is granted through the Copyright Clearance Center. 27 Congress Street. Salem, Massachusetts 01970, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of PubLications. which contains an alphabetical title list (with full ordering information) and indexes of subjects, authors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher. Department F, The World Bank, 1818 H Street, N.W.. Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'lena, 75116 Paris, France. Library of Congress Catalogilng-in-Publicatilon Data Skillicorn, Paul, 1950- Duckweed aquaculture : a new aquatic farming system for developing countries / Paul Skillicorn, William Spira, and William Journey. p. cm. Includes bibliographical references. ISBN 0-8213-2067-X 1. Duckweeds. 2. Duckweeds-Utilization. 3. Duckweeds- -Developing countries. I. Spira, William, 1946- . II. Journey, William. 1943- . III. International Bank for Reconstruction and Development. IV. Title. SB317.D82S38 1993 639'.89-dc2O 92-45127 CIP Table of Contents Foreword ...................................... vii Preface ...................................... ix Section 1 - Biology of Duckweed .......................................1 Morphology .......................................1 Distribution .......................................1 Growth conditions ...................................... 2 Production rates ...................................... 3 Nutritional value ....................................... 4 Section 2 - Duckweed Farming ....................................... 8 Land ....................................... 8 Water management ...................................... 9 Nutrient sources ...................................... 9 Nitrogen ...................................... 10 Phosphorus ...................................... 11 Potassium ...................................... 11 Trace minerals ...................................... 12 Organic wastes ...................................... 12 Fertilizer application ...................................... 12 Crop management ...................................... 13 Containment and wind buffering .................................... 14 Seeding duckweed ...................................... 17 Stress management ....................................... 17 Unicellular algae ....................................... 18 Harvesting ....................................... 18 Section 3 - Duckweed-Fed Fish Production ........................ 22 Introduction ...................................... 22 Importance of oxygen ...................................... 23 More efficient culture of top-feeders ............................... 23 Review of conventional carp polyculture ............................ 23 Fertilization ...................................... 24 Supplementary feeding ...................................... 25 Production constraints ...................................... 25 Typical carp yields in Asia ...................................... 26 Duckweed-fed carp polyculture ....................................... 26 Practical objectives ....................................... 26 Logic of duckweed-fed carp polyculture .......................... 27 iil Al Zg -............. . tqqsjjod fjleAajseM paaLxna 1 9 siueild poava.np pajs)A-q jo 4aSes 19 .Aua;opa ajeldn jUaTfnN 09 ........o........pVlls aA. 89.BUns3aseH iS. slueld paoaMAnp ?uTuTejuoa pue iuflnql.sTQ LS .StulsAs Vuf-eJna3TaaH 99 ..................................eTIj paam3mf 99. .uniss A0oU ?nd pasaMXznu 99.Lo,I-j.Iu0- -T......... eodpo@Pl 99 ................................ esodsjp a~pniS strS.unjXss XnwwIFd 6t'.m~~~~~~~~....uo; uuaig om.z 61 . ----.............................................. 9m oXj9f$ ,u* l amXs, 1oUAftolus posgs-pooaaong - 9 U°l402s t. sanssl iuIlz)d 9t .suawarinbai IlpT.3 St . ....sanss1 uolsuaJxa pue awO1Wssse I3utqlJaL bb 7s;sAp3wo fe¶uFr1 bbthO JoFIeJLu WeaPaA .b W tsapeUF dnoa .b ae3ufW 4Ufl-OA¶J .besppoul pueuzag .buogonposd i4s§^ pue p~a*ipnp Jo aThnxqF it'e'u -. jvS uopn;}uu pUB o;mouoox - p uo;S 8£.eldeff paJ-p3aMpflQ S£99.W@6t@se@X uT-inomFOt u~axho pue dos3 am oi p;xedtuIo AjipFpanpoad pue jusuxaBeuey ££9.3WpJ uafAxo 99. puod atf JO UORzFIIfA, .£psaj p33MA3{ofl .Z als upaoXs dser .Z3mnXlnoAJod dno p35-psava&np inoqe sasatplociLq ajseq Pathogen removal ................................................... 62 Final effluent discharge .................................................. 63 Commercial systems ................................................... 63 Section 6 - Alternative Uses for Duckweed, Constraints and Future Research ................................................... 65 Developing alternative uses for duckweed .......................... 65 Duckweed as poultry and animal feed ............................ 65 Duckweed as a mineral sink .......................................... 66 Constraints and research needs ........................................ 66 Duckweed production ................................................... 67 Genetic improvement ................................................... 67 Duckweed wastewater treatment .................................... 67 Drying ................................................... 68 Derived products ................................................... 68 Duckweed and fisheries ................................................. 68 Annexes Investment Scenarios ................................................... 69 Annex 1 Investment Scenario for Duckweed-Fed Fish Production ................................................... 70 Annex 2 Investment Scenario for Duckweed Production ..... 71 Selected Bibliography Duckweed ................................................... 72 Fish culture ................................................... 75 Figures 1 Duckweed, the smallest flowering plants .2 2 Composition of duckweed from three sources .4 3 Comparison of lysine and methionine content of protein from various sources .5 4 Pigment content of several samples of duckweed growing wild on wastewater .6 5 Protein content of various animal feedstuff ingredients .7 6 Making a duckweed culture pond .8 7 Protecting duckweed from wind and wave action .8 8 Nutrients for duckweed can come from fertilizer or organic wastes .10 9 Co-cropping with terrestrial plants mimics duckweed's natural environment and increases cropping intensity .15 10 Collecting duckweed seedstock .16 v 11 Growth in excess of the optimal stocking density should be harvested regularly to promote rapid growth ............ .......................... 19 12 Harvesting by skimming with a dip net ................... 20 13 Drying duckweed in the sun and bagging dried meal in opaque plastic bags ......................... 21 14 Chinese and Indian carp species ............................ 22 15 Fish inputs (1989-90) ............................. 30 16 Duckweed inputs (1989-90) ............................. 30 17 Fresh duckweed from the culture pond is fed directly to carp in the fish pond ........................... 31 18 Weight of fish caught (1990) ........................... 31 19 Average weight of fish catch by month in Mirzapur duckweed-fed carp production tests (1990) ............ 32 20 Average weight of fish catch after 13 months .......... 33 21 Market-size fish are selected and weighed ............... 37 22 Major tilapia species ...................................... 38 23 Product flows in integrated farming of duckweed, fish and poultry .43 24 Model duckweed wastewater treatment system using floating containment barriers .59 25 Model duckweed wastewater treatment system using earthen berms for crop containment . ........... 60 Tables 1 Daily Fertilizer Application Matrix ............................ 13 2 Quality of final treated effluent for March 23, 1991 ......................... 52 Box 1 Box 1. Wastewater Treatment ......................... 50 vi Foreword Although duckweed species are familiar to most people who have seen the tiny aquatic plants covering stagnant water bodies, few people realize their potential. Until a few years ago, man made little use of duckweed species. Their unique properties, such as their phenomenal growth rate, high protein content, ability to clean wastewater and thrive in fresh as well as brackish water, were only recognized by a few scientists. Prior to 1988 duckweed had been used only in commercial ap- plications to treat wastewater in North America. In 1989 staff of a non-governmental organization based in Columbia, Maryland, The PRISM Group, initiated a pilot project in Bangladesh to develop farming systems for duckweed and to test its value as a fish feed. An earlier project in Peru invesugated the nutritional value of dried duckweed meal in poultry rations. The results of the pilot operations were extremely promising; production of duckweed-fed carp far exceeded expectations, and dried duckweed meal provided an excellent substitute for soy and fish meals in poultry feeds. Duckweed could be grown using waste- water for nutrients, or alternatively using commercial fertilizers. During start-up of the pilot operations it also became apparent how little is known about the agronomic aspects of producing var- ious species of the duckweed family, and exactly why it is so effec- tive as a single nutritional input for carp and other fish. Although these pilot operations were located in South Asia and Latin America, the results suggested that the plant would be im- portant as a source of fish and poultry feed and simultaneously as a wastewater treatment process in selected areas of the Middle East, particularly in Egypt and Pakistan. Technical and agronomic information about duckweed culture and feed use, and details of farming duckweed and fish in a single system, are not easily available to the general public, let alone to fish farmers in developing countries. The pilot operations in Bangladesh demonstrated that duckweed and fish culture can succeed commer- cially, although such ventures would initially require technical as- sistance and information. In many other areas of the world pilot vii operations linked to applied research may be required to review pro- duction parameters before commercial operations should be initiat- ed. This Technical Study was therefore designed to bring together, in one publication, relevant information on duckweed culture and its uses to make people worldwide aware of the potential of this plant, to disseminate the currently available technical and agronomic in- formation, and to list those aspects that require further research, such as duckweed agronomy, genetics and use in animal feeds. This Technical Study is aimed at the following audiences: (a) established fish farmers who would like to experiment with duck- weed as a fish feed, and staff of agricultural extension services in- volved in fish culture; (b) scientists of aquaculture research institutes who may initiate pilot operations and applied research on duckweed; (c) staff of bilateral and multilateral donor agencies who may promote funding for duckweed research and pilot opera- tions; and (d) wastewater specialists in governments and donor agencies who may promote wastewater treatment plants based on duckweed in conjunction with fish culture. The information in this technical study comes from many sources; the contribution of the staff of the Mirzapur experimental station in Bangladesh and its director Mohammed lkramullah, in particular, is acknowledged. Paul Skillicorn and William Spira of the PRISM Group, and William Journey wrote the text. Viet Ngo of the Lemna Corporation and Richard Middleton of Kalbermatten As- sociates provided technical material relating to wastewater treat- ment applications. The draft was reviewed by a Bank technical committee comprising Messrs. Grimshaw, Khouri, Leeuwrik, van Santen and Macoun. Professor Thomas Popma of the International Center for Aquaculture at Auburn University provided technical support, Professor Guy Alaerts of the International Institute for Hy- draulic and Environmental Engineering Delft, the Netherlands, re- viewed the section on Wastewater Treatment, and illustrations were provided by Ms. S. Gray of Auburn. Harinder S. Kohli Director, Technical Department Europe, Middle East and North Africa Region vili Preface The purpose of this booklet is to present a group of tiny aquatic plants commonly known as "duckweeds" as a promising new com- mercial aquaculture crop. Duckweed species are members of the taxonomic family Lemnaceae. They are ubiquitous, hardy, and grow rapidly if their needs are met through sound crop manage- ment. Aquaculture systems are many times more productive than terrestrial agriculture and have the potential to increase protein production at rates similar to increases of terrestrial carbohydrate crops realized during the Green Revolution. Section 1 presents basic information on duckweed biology. This paper summarizes current knowledge, gained from prac- tical experience from the beginning of 1989 to mid-1991 in an ex- perimental program in Mirzapur, Bangladesh, where duckweed cultivation was established and fresh duckweed fed to carp and tilapia. In the Mirzapur experimental program a farming system was developed which can sustain dry-weight yields of 13 - 38 met- ric tons per hectare per year (ton/ha/year). which is a rate exceed- ing single-crop soybean production six to tenfold. Section 2 discusses duckweed farming issues in detail. Like most aquatic plants, duckweed species have a high wa- ter content, but their solid fraction has about the same quantity and quality of protein as soybean meal. Fresh duckweed plants appear to be a complete nutritional package for carp and tilapia. Duckweed-fed fish production does not depend on mechanical aeration and appears to be significantly more productive and eas- ier to manage than traditional pond fish culture processes. Sec- tion 3 addresses the important issues in duckweed-fed fish production. The economics of duckweed farming and duckweed-fed fish production and institutional factors that are likely to affect its wide- spread adoption as a commercial crop are discussed in Section 4. Section 5 summarizes the use of duckweed for stripping nu- trients from wastewater. The bio-accumulation of nutrients and dissolved solids by duckweed is highly effective. World-wide appli- cations of duckweed-based technologies for wastewater treatment ix and re-use are being implemented in both idustrialized and devel- oping countries. Section 6 provides other potential commercial applications of duckweed: (1) in its dried form as the high protein component of animal feeds, and (2) as a saline-tolerant aquaculture crop. It also contains a discussion of key research issues and constraints inhib- iting the potential for duckweed as a commercial crop. The paper concludes with a selected bibliography covering im- portant duckweed-related research. This is an impressive body of literature covering the entire spectrum from microbiology to poultry research. The work described here did not attempt to repeat exper- imentation of earlier researchers, nor did it originate any basic duckweed production or application concepts. The concepts pre- sented here do, however, represent the first attempt to synthesize a complete commercial paradigm for cultivating and using duckweed. x Section 1 - Biology of Duckweed Duckweed species are small floating aquatic plants found world- wide and often seen growing in thick, blanket-like mats on still, nutri- ent-rich fresh and brackish waters. They are monocotyledons belonging to the botanical family Lernnaceae and are classified as high- er plants, or macrophytes, although they are often mistaken for algae. The family consists of four genera, Lenumn, Spirodela, Woiffia, and Wolf- fiella, among which about 40 species have been identified so far. All species occasionally produce tiny, almost invisible flowers and seeds. but what triggers flowering is unknown. Many species of duckweed cope with low temperatures by forming a special starchy "survival" frond known as a turion. With cold weather, the turion forms and sinks to the bottom of the pond where it remains dormant until rising temperatures in the spring trigger resumption of normal growth. Morphology Duckweed species are the smallest of all flowering plants. Their structural and functional features have been simplified by natural selection to only those necessary to survive in an aquatic environment. An individual duckweed frond has no leaf, stem, or spe- cialized structures; the entire plant consists of a flat, ovoid frond as shown in figure 1. Many species may have hair-like rootlets which function as stability organs. Species of the genus Spirodela have the largest fronds, measuring as much as 20 mm across, while those of Wo!fia species are 2 mm or less in diameter. Lemna species are intermediate size at 6 - 8 mm. Compared with most plants, duckweed fronds have little fiber- as lit- tle as 5 percent in cultured plants-because they do not need struc- tural tissue to support leaves or stems. As a result virtually all tissue is metabolically active and useful as a feed or food product. This im- portant characteristic contrasts favorably with many terrestrial crops such as soybeans. rice, or maize, most of whose total biomass is left behind after the useful parts have been harvested. Distribution Duckweed species are adapted to a wide variety of geographic and climatic zones and can be found in all but waterless deserts and permanently frozen polar regions. Most, however, are found in moderate climates of tropical and temperate zones. Many 1 D Figure 1. Duckweed, the smallest flowering plants Genera: A. Spirodela B. Lemma C. Wolffia D. Wolfiella E. Lemna with Wolffia species can survive temperature extremes, but grow fastest under warm, sunny conditions. They are spread by floods and aquatic birds. Duckweed species have an inherent capability to exploit favora- ble ecological conditions by growing extremely rapidly. Their wide ge- ographic distribution indicates a high probability of ample genetic diversity and good potential to improve their agronomic characteris- tics through selective breeding. Native species are almost always avail- able and can be collected and cultivated where water is available, including moderately saline environments. Growth conditions The natural habitat of duckweed is floating freely on the surface of fresh or brackish water sheltered from wind and wave action by surrounding vegetation. The most favorable cir- cumstance is water with decaying organic material to provide duck- weed with a steady supply of growth nutrients and trace elements. A dense cover of duckweed shuts out light and inhibits competing sub- merged aquatic plants. including algae. Duckweed fronds are not anchored in soil, but float freely on the surface of a body of water. They can be dispersed by fast currents or 2 pushed toward a bank by wind and wave action. If the plants become piled up in deep layers the lowest layer will be cut off from light and will eventually die. Plants pushed from the water onto a bank will also dry out and die. Disruption of the complete cover on the water's sur- face permits the growth of algae and other submerged plants that can become dominant and inhibit further growth of a duckweed colony. To cultivate duckweed a farmer needs to organize and main- tain conditions that mimic the natural environmental niche of duckweed: a sheltered, pond-like culture plot and a constant sup- ply of water and nutrients from organic or mineral fertilizers. Wastewater effluent rich in organic material is a particularly valu- able asset for cultivating duckweed because it provides a steady supply of essential nutrients and water. In this case there is a coincidence of interests between a mu- nicipal government, which would treat the wastewater if it could af- ford to do so, and nearby farmers, who can profitably do so. Production rates Duckweed reproduction is primarily vegeta- tive. Daughter fronds bud from reproductive pockets on the side of a mature frond. An individual frond may produce as many as 10 generations of progeny over a period of 10 days to several weeks be- fore dying. As the frond ages its fiber and mineral content increas- es, and it reproduces at a slower rate. Duckweed plants can double their mass in less than two days under ideal conditions of nutrient availability, sunlight, and tem- perature. This is faster than almost any other higher plant. Under experimental conditions their production rate can approach an ex- trapolated yield of four metric tons/ha/day of fresh plant biomass, or about 80 metric tons/ha/year of solid material. This pattern more closely resembles the exponential growth of unicellular algae than that of higher plants and denotes an unusually high biological potential. Average growth rates of unmanaged colonies of duckweed will be reduced by a variety of stresses: nutrient scarcity or imbalance: tox- ins: extremes of pH and temperature; crowding by overgrowth of the colony: and competition from other plants for light and nutrients. Actual yields of fresh material from commercial-scale culti- vation of Spirodela, Lemna, and Woiffla species at the Mirzapur 3 experimental site in Bangladesh range from 0.5 to 1.5 metric tons/ha/day. which is equivalent to 13 to 38 metric tons/ha/ year of solid material. Nutritional value Fresh duckweed fronds contain 92 to 94 percent water. Fiber and ash content is higher and protein con- tent lower in duckweed colonies with slow growth. The solid frac- tion of a wild colony of duckweed growing on nutrient-poor water typically ranges from 15 to 25 percent protein and from 15 to 30 percent fiber. Duckweed grown under ideal conditions and har- vested regularly will have a fiber content of 5 to 15 percent and a protein content of 35 to 45 percent, depending on the species involved, as illustrated in figure 2. Data were obtained from per cent 55 - 50 - Legend: Z4 Lagoon Inlet 45 - Inlet + 50 m 40 - Enriched Culture 35- 25- 20- 15 10 Protein Fiber Ash Fat Figure 2. Composition of duckweed from three sources Source: Mbagwu and Adeniji, 1988 4 duckweed colonies growing on a wastewater treatment lagoon and from a duckweed culture enriched with fertilizer. Duckweed protein has higher concentrations of the essen- tial amino acids, lysine and methionine, than most plant pro- teins and more closely resembles animal protein in that respect. Figure 3 compares the lysine and methionine concentrations of proteins from several sources with the FAO standard recom- mended for human nutrition. Cultured duckweed also has high concentrations of trace minerals and pigments, particularly beta carotene and xantho- phyll, that make duckweed meal an especially valuable supple- ment for poultry and other animal feeds. The total content of carotenoids in duckweed meal is 10 times higher than that in Legend: FAO reference -ethioni_ // / -- .i j, // _ ~Methionine U Lysine Cottonseed meal Groundnut meal Soybean meal / Duckweed meal - Blood meal 7, / /' 0 1 2 3 4 5 6 7 8 9 10 11 grams/ 100 grams Figure 3. Comparison of lysine and methionine content of protein from various sources Source: Agagwu and Adeniji. 1988 5 terrestrial plants; xanthophyll concentrations of over 1,000 parts per million (ppm) were documented in poultry feeding trials in Peru and are shown in figure 4. This is economically important because of the relatively high cost of the pigment supplement in poultry feed. A monoculture of Nile tilapia and a polyculture of Chinese and Indian carp species were observed to feed readily on fresh duckweed in the Mirzapur experimental program. Utilizing duckweed in its fresh, green state as a fish feed minimizes han- Legend: L. minor K PLANKTONFEED -> FISH is more efficient with respect to inputs than: DUCKWEEDpEED -> FISH which would indeed be the case, if there were no oxygen con- straint. Considering oxygen as a constraint, however, it is useful to extend the model as follows: !OXYGENvA,VAJ FERTILIZERCHnM -> PLANKTONFEED -> FISH -> FECES- FERT -> NH3, PLANKTONFEED -> FISH -> FECESFERT- -> NH3, PLANKTONFEED ...OXYGENArJl is less efficient with respects to inputs and oxygen than: duckweedfeed ->ftsh ->fecesfeed ->fish -> Ioxygenava.lJfecesfe,t -> planktorle)ed ->fish ->fecesfert -> NH3, planktonfeed ->fish ->fecesfert -> NH3, planktonfed -> .oxygenmin] 27 In the duckweed model, the entire cycle of: duckweedfeed ->fish ->fecesjed ->fsh -> takes place ahead of the existing oxygen constraint. The second round of fecal input from bottom-feeding carp is then roughly anal- ogous to the chemical or organic fertilizer input to conventional carp polyculture, but at a lower level. The fish farmer must, of course, balance this potential in- crease in productivity against his increased costs. Technological in- puts in the duckweed model do not differ from conventional non- aerated carp polyculture. The additional cost of a duckweed system is, therefore, roughly equal to the price of duckweed inputs. A more careful analysis should also consider increased in- cremental costs for fingerling inputs, as well as decreased ex- penses for fertilizer and manure, which a farmer would otherwise expect to incur following conventional carp polyculture method- ology. For simplicity, however, unadjusted duckweed procure- ment costs are used to estimate the cost of converting to a duckweed polyculture system. Because the feces of top-feeders and first-round bottom-feed- ers provide the manure normally purchased to meet the needs of middle and second-round bottom-feeders, the farmer has only to calculate the profit for the incremental production of top-feeders (grass carp, catla, and mirror carp) and bottom-feeders (mrigal and mirror carp) to determine his marginal benefit. Experience in the Mirzapur experimental program in Bangla- desh has been that a grass carp/mrigal combination produces 1 kg of fish for between 10 to 12 kg of fresh duckweed, or about $0.30 to $.401 worth of duckweed consumed. That amount of fish brought approximately $1.50 at the wholesale price. The farmer is, in effect, making a large profit on his "fertilizer production engine". Carp stocking strategy In the Mirzapur experimental ponds, grass carp (Ctenopharyngodon idella) is the primary consumer of duckweed in the polyculture. However, both catla (Catla catla) and mirror carp (Cyprinus carpio) also compete aggressively for 1AII dollar amounts are US$ 28 available duckweed feed and consume it directly. Top-feeders di- rectly absorb about 50 percent of duckweed nutrients in their di- gestive systems. Their feces contain the balance of the original duckweed nutrients and furnish a relatively high quality detritus for bottom-feeders. Bottom-feeding species comprise a relatively high 30 percent of the polyculture. The purpose is to increase the probability that fec- es from the entire fish population will be digested several times, not only to convert the maximum amount of nutrients into fish flesh, but to moderate biochemical oxygen demand in the pond. Mrigal (Cirrhmnus mrigala) is a bottom-feeder and is tolerant of the low ox- ygen levels at the bottom. Although they grow more slowly than the other varieties, they keep the pond bottom clean. Rohu (Labeo rohita) and silver carp (Hypothalmichthys molitri are two phytoplankton-feeding species used in the duckweed-fed polyculture at a total of 40 percent of the species mix, or approxi- mately half of the typical Chinese carp polyculture. The objective in the Mirzapur experimental program was to match the fish popula- tion to the expected lower availability of phytoplankton. Maintain- ing a proper balance between middle-feeders and phytoplankton production achieves a higher efficiency in fish flesh production and reduces fluctuations in dissolved oxygen caused by excessive den- sities of green algae. Carp fry and fingerlings feed on zooplankton. Fingerlings will also eat Wolffia as soon as their mouths are big enough. The tradi- tional use of duckweed in Asia has been to feed fish fingerlings. Production data shown in figures 15, 16, 18, 19, and 20, refer to the first 12 months (of an 18 month cycle) of carp polyculture production at the Mirzapur experimental carp pond, a 2.2 hectare pond stocked with approximately 50,000 carp in September 1989. As of April 1991 approximately 18 tons of the original fish had been harvested. An estimated three to five tons, primarily mirror carp, were stolen, and an estimated five tons of the original fish were left in the pond. A further 30,000 fingerlings were stocked in the pond in September 1990. Harvesting of these fish, along with the remain- ing original fish, began in April 1991. Although total pond produc- tivity can only be estimated, it appears to be around 10 tons per hectare per year. 29 Mirzapur Duckweed-Fed Carp Production Fingerling Inputs (N = 55,000) Catla Carp (15 %) Rohu Carp (15 %) Mrigal Carp (20 %) Grass Carp (20 %) Silver Carp (20 %) Mirror Carp(1 0 %) Figure 15. Fish inputs (1989-90) Cumulative Duckweed Production 120 100 80 0 -60 E 40 20 0 A S O N D J F MA M J J A S month Figure 16. Duckweed inputs (1989-90) 30 Figure 17. Fresh duckweed from the culture pond is fed directly to carp in the fish pond 14 12 ; / Mrigal carp c 10 Catla carp 0) ., 8 Rohu carp s 6 Grlass lcarp M A M J J A S 0 month Figure 18. Weight of fish caught (1990) 31 Duckweedfeed Duckweed is not a supplementary feed in the Mirzapur polyculture, it is the main source of nutrition. Feeding a carp polyculture with duckweed simplifies nutrition to a single input and the feeding schedule to a single issue: feeding the carp as much as they will eat. Any uneaten duckweed will be visible floating in the feeding station and the farner can respond by reduc- ing the volume on the following day. Fish are fed duckweed throughout the day. Freshly harvested duckweed is brought in baskets to the pond and distributed evenly among several "feeding stations" consisting of 4 m2 open-bottom enclosures, as illustrated in figure 17. Feeding stations provide ac- cess by the fish to the duckweed and prevent it from dispersing over the pond surface. The feeding station can be a floating enclosure anchored near the shore. Six feeding stations per hectare were Mirzapur Duckweed-Fed Carp Production Average Weight of Catch by Month Silver Carp Mirror Carp Grass Carp ____ 1 .5- ____ ____ % Rohu Carp I--< \ w _ Catla Carp Mrigal Carp M A M J J A S O month Figure 19. Average weight of fish catch by month In Mirzapur duckweed-fed carp production tests (1990) 32 4 3.5 - 3- 2.5 _ 2 0.5 Silver Mirror Grass Rohu Catia Mrlgal carp species Figure 20. Average weight of fish catch after 13 months installed at the Mirzapur experimental site and appeared to provide sufficient access to food for all fish. Judging from carp production rates in the Mirzapur experi- mental program, approximately 10 to 12 kg of fresh, cultured duck- weed is converted into 1 kg of fish. Precise confirmation of this figure awaits controlled experimentation. Fertilization of the pond Fertilization of a duckweed-fed fish culture is indirect and gradual, resulting from bacterial decompo- sition of fish feces, dead algae, and other fermenting organic mate- rial in the pond. The issue of pond fertility is removed from the farmer's management tasks. Fertilization of the base of the food web in the flsh pond is automatically regulated by the consumption of fresh duckweed by the fish and its subsequent entry into the pond water where it will ultimately decompose. Oxygen regime In the Mirzapur experimental model, several carp species acquire a significant percentage of their nutritional re- quirements through direct consumption of duckweed. This allows maintenance of higher stocking densities while also reducing 33 production of algae that contributes to depletion of oxygen during nocturnal respiration. The result is a pond environment that has generally higher concentrations of dissolved oxygen with a lower amplitude of diurnal oxygen fluctuation. This means more fish, healthier fish, and more confident farmers. The dawn-dissolved oxygen concentration in a 0.5 ha pond at the Mirzapur experimental site, stocked with 30,000 fish fed en- tirely on duckweed, was monitored over a six-month period. It did not go below 4 milligrams per liter (mg/i) until the fish density in- creased to an estimated 20 metric tons/ha and the temperature began to rise with the advent of spring in Bangladesh. Feeding was curtailed to reduce pond BOD, and the stock of fish was re- duced by harvesting until only about 15 metric tons/ha remained. This again prevented dawn-dissolved oxygen levels from dropping below 4 mg/l. Management and productivity compared to the tradition- al Chinese model The Mirzapur duckweed-fed carp polyculture model has an 18-month cycle. Fingerlings are introduced in August and September, harvesting begins in March and continues for ap- proximately one year. A second 18-month cycle begins the following year and continues concurrently for six months. After the initial six months, the model allows year-round harvesting. In the Mirzapur experimental system, duckweed is the single nutrient input. It floats and is visible until eaten. This minimizes ambiguities concerning the level of feeding needed to support effi- cient fish growth. Fish regulate their feeding by eating until they are satiated. The farmer has a simple visual signal to regulate the feed supply and will supply just enough to guarantee a small daily residual floating in the feeding station. Over-feeding and over-ferti- lization are two problems typical of the traditional model which are avoided in the duckweed-fed polyculture. However, for this model to be risk-free it is essential that optimal stocking rates be known precisely, which is not yet the case. Duckweed species grow faster in warm weather when fish need more feed and more slowly in cold weather when the fish also do not require as much feed. In general a farner should design a duckweed supply capability to fulfill his peak needs and should dry excess biomass for use as an animal feed ingredient. Current 34 production rates suggest that one hectare of duckweed production can support two hectares of carp polyculture. The first annual cycle of carp production in Bangladesh pro- duced slightly more than 10 metric tons/ha/year. This yield oc- curred in spite of the fact that, for the first three months, duckweed production constraints prevented the fish from receiving enough duckweed feed for optimal growth. Empirical results so far in Bangladesh suggest that a polycul- ture stocked at about 30,000 fish per hectare may be fed as much duckweed as they will eat daily, regardless of the season. Further- more, a yield of between 10 to 15 tons/ha/year appears to be sus- tainable before biological constraints become the limiting factors. The Mirzapur duckweed-fed fish polyculture requires daily la- bor over the entire season. Carp are fed daily and duckweed is har- vested daily to maintain the best production rates. The duckweed farmer's family is the most cost-effective source of labor and can be gainfully employed year-round. Hired labor is usually necessary at critical times, such as weekly harvests and pond-cleaning. Crop and oxygen monitoring Unicellular algae. or phyto- plankton, grow extremely rapidly in response to nutrient availabil- ity, sunlight, and warm temperatures. These algae are harvested for food by filter-feeding species of carp and other phytoplankton- feeding fish. An oversupply of phytoplankton can deplete the dis- solved oxygen in the pond to dangerously low levels for the fish. Sudden die-off of phytoplankton and its subsequent decay results in a dramatic increase in BOD that can also deplete oxygen to dan- gerously low levels. Direct monitoring of pond-dissolved oxygen levels is impracti- cal for most small farmers in countries such as Bangladesh. Equip- ment is too expensive to enable widespread use and not sufficiently robust for continuous use. However, monitoring of pond oxygen can be performed during harvesting. Fish with adequate oxygen ex- hibit considerable vigor during harvesting. When oxygen levels fall below 4 mg/l the reduction of jumping during harvesting is strik- ing. If farmers harvest twice a week, observation of fish behavior during harvesting should provide feedback in time to reduce feed inputs, to introduce fresh water, or to further reduce stock, all of which can have immediate impact on pond-dissolved oxygen levels. 35 Fish quality, health and security Duckweed-fed carp raised in the Mirzapur experimental program have so far appeared to be healthy and well-nourished. However, the bottom-feeding mrigal, the slowest growing of the species in the polyculture, averaged 0.45 kg in one year of growth. In this duckweed-fed system mrigal feed primarily on detritus provided by the fecal matter of the top-feed- ers. which has only a fraction of the nutrients of fresh duckweed. The relatively poor production of mrigal is attributable to the strat- egy of stocking them in relatively high numbers so that fecal matter from top-feeders would be more likely to be consumed before con- tributing to pond BOD. Figure 20 demonstrates the average weight of fish caught 13 months after being placed in the Mirzapur experimental pond. Sil- ver carp experienced the best growth at 2.75 kg/year, followed by catla and rohu. The relatively poor growth of grass carp attests to their high stocking density and a shortage of duckweed during the first several months of production. Grass carp production during the second production cycle (not reported here), when duckweed inputs were not constrained, was considerably higher with individ- ual fish reaching 4 kg within six months. Mirror carp growth was, in fact, better than indicated. Only a few, stunted mirror carp remained in the pond at the end of one year. Mirror carp are easily caught from the pond perimeter by throw net, and most were stolen by intruders before action was tak- en to increase pond security. Once the value of the fish in the Mir- zapur experimental ponds became known, it became necessary to employ nighttime guards. Management of the security force is an added concern and operating cost. Fish mortality has not been an issue so far in the Mirzapur exper- imental program. There have been no fish kills or outbreaks of dis- ease. Water quality appears to be good and the fish appear to be in good health, even at relatively high densities for the semi-intensive system. Harvesting Regular and frequent harvests are prescribed for duckweed-fed fish culture. The catch is sorted by size, counted, and weighed. The intermediate size fish are returned to the pond for further growth. These data help the farmer to track the 36 growth rate of his fish and to estimate the quantity and quality of future harvests. Routine harvesting of duckweed-fed carp began approximately six months after the Mirzapur polyculture pond was stocked. Bi- weekly harvesting was the preferred pattern, following a simple protocol to take the largest fish (75 to 100 percentile) and the small- est (O to 25 percentile) in each species. The rationale is the assump- tion that the largest fish will have a declining growth rate and that the small fish are simply poor performers. This protocol was partic- ularly difficult to follow with respect to mrigal which, because of their small size, became entangled in the nets. Fish damaged in this manner were removed from the pond regardless of size. As the carp were harvested, they were counted, each variety weighed separately, and the data recorded in order to analyze the efficiency of the farming operation and to maintain the desired ra- tios of species in the pond. This is illustrated in figure 21. Care was taken not to deplete top-feeders and bottom-feeders- the fertilizer and food engines-disproportionately. Fortunately, several species of carp, not considered to be macrophyte feeders Figure 21. Market-size fish are selected and weighed 37 (mirror carp and catla, in particular), also competed vigorously for supplies of fresh duckweed, which is apparently a learned behavior. Markets Duckweed-fed fish from the Mirzapur experimental site had a clear quality edge in the local market. Aesthetically, fresh, green duckweed contrasted favorably with manure and other less appealing inputs to a conventional pond fishery. The consum- er's perception appeared to be that because duckweed-fed fish are reared on fresh vegetables and live in higher quality water, they "smell, feel, and taste" better than fish reared conventionally. Duckweed-fed tilapia Tilapia species are of African origin but have been introduced to most tropical and subtropical regions. (See figure 22). Tilapia are hardy, grow fast, and can tolerate low pond oxygen levels better than most fish. They are warm water fish which o. niIO+iCq C). Liv,rec1. o' 0. Maossclmbrca Figure 22. Major tilapia species 38 do not grow below 16° C and do not survive temperatures below 100 C. Unlike carp, they have no "floating" intramuscular bones, mak- ing it easier for the diner to separate bones from flesh. Most species of tilapia tolerate brackish water well. Adult tila- pia are primarily herbivorous, occasionally omnivorous, and some species are used to control aquatic weeds. Fry feed primarily on plankton. At least one species, Oreochromis niloticus, is reported to be extremely flexible in its feeding habits, readily consuming Lemna and Wotffla species along with phytoplankton and detritus. Tilapia are well-equipped to feed on duckweed. They have grinding plates in their pharynx, a highly acidic stomach, and a long intestine to absorb digested nutrients. Duckweed supplies the high protein diet they need for rapid growth. The maceration and digestion of duckweed by macrophyte-feeding tilapia requires less energy expenditure than a diet of more fibrous plants. Because the Nile tilapia appears to be able to harvest food from all of the space and food niches in a pond, it was tested in the Mirzapur experimental program as an alternative to the duck- weed-fed carp polyculture. The single-species culture appears to benefit from duckweed as the single nutritional input in much the same way as the carp polyculture because the nutrients appear to be distributed similarly. Production at Mirzapur in a 0.6 hectare pond totaled 4.5 tons in one year of continuous operation. As management of the pond improved, and the stocking balance be- tween recruits, juveniles, and mature fish became more efficient, productivity rates improved. Local pond managers now believe that they should be able to average at least 10 tons/ha/year for mixed (sizes) tilapia harvest. Because of their fecundity, tilapia require special management to keep their population stable and to maintain even growth. They mature at about three months and breed prolifically in the pond at intervals of three to six weeks. The additional fish population. called recruits, leads quickly to extreme competition for food and, hence, a stunted population. There are four basic approaches to management of tilapia populations: monosex culture, intensive culling, production in brackish water and inclusion of predators. Frequent, intensive harvesting to remove market-size fish and re- cruits is highly labor intensive and can stress the fish population. 39 It is, however, a relatively simple technique available to the small farmer. Predatory fish can be included with the tilapia culture to con- trol recruits and allow the production of market-size fish. Predator species include the clarias catfish, notopterus, snakehead, and others, many of which have high market value. The principle con- straints with this method are the difficulty of obtaining stocks of predator species and determining efficient stocking densities. The tilapia culture strategy investigated at the Mirzapur exper- imental site is conceptually similar to duckweed cultivation. The concept is to determine an efficient "standing crop" and to maintain it with bi-weekly harvests. Tilapia are categorized either as recruits, adolescents, or adults. During harvests, estimates are made of the total amount of tilapia in the pond and their distribution among the three categories. For example, the standing crop today is 10 tons and, numerically, 60 percent of the fish are recruits, 30 percent are adolescents, and 10% are adults. To bring the standing crop back to the empirically derived "normative" size and balance, the har- vesting heuristic should then specify a harvest profile by weight: harvest 400 kg- 50 kg of recruits, 150 kg of adolescents, and 200 kg of adults. Current harvest profiles will rely more on intuition than formula until efficient harvesting algorithms are developed. Tilapia recruits, although very small, fetch a surprisingly high market price in rural markets in Bangladesh. They are purchased by people unable to afford fish in the size range prevalent in the market (0.5 - 1 kg). Where tilapia above 250 g can command up to $2.00 per kg in rural markets, mixed adolescents, and recruits can bring up to $1.00 per kilogram. This mechanism allows even the poorest people to include some fish in their diet. With production costs averaging between $0.40- $0.50 per kg in Bangladesh, farm- ing duckweed-fed tilapia is highly profitable. 40 Section 4 - Economic and Institutional Issues Introduction of duckweed cropping is likely to be attended with "teething problems" influenced by several factors in unfamiliar combinations. Duckweed is not only a novel crop, but a highly in- tensive one. It appears to be "multipurpose" in the sense that it may be farmed in several possible settings with different economic and financial implications, and it is an aquaculture crop. With the ex- ception of the Mirzapur experimental program, there are no at- tempts on record to develop full-scale cropping systems. There are currently no institutions equipped to provide extension support to duckweed farmers, and a market for duckweed does not yet exist. Nevertheless, the success of the experimental work suggests that duckweed cropping should be introduced to a wider audience of farmers, especially those in tropical and semitropical developing countries. A logical first step would be to develop institutional centers ca- pable of assimilating existing knowledge concerning duckweed, adapting this knowledge to specific local conditions and expanding it through research. These research and demonstration centers should also be supported by extension and credit institutions ca- pable of delivering information and financial support directly to duckweed and duckweed-fish farmers. Pending the development of markets for duckweed as an end-product, mechanisms should be developed to link duckweed production with some end-use. Cur- rently there are only three: direct fish or poultry production, and production of blended animal feeds. The remainder of this section will discuss key institutional is- sues at the farm level and beyond, which should be addressed to facilitate introduction of duckweed production and duckweed-fed fish production elsewhere. The research center model, best exem- plified by the various CGLAR facilities worldwide, needs little elab- oration. The discussion will concentrate, therefore, on farm level linkages, extension, credit, and pricing issues that are basic to duckweed production. 41 Linkage of duckweed and fish production Duckweed cannot be stored for more than two or three days in its green state and at temperatures above 200 C. Until adequate cold storage or drying technologies have been developed, this limitation prevents forma- tion of a conventional duckweed market where supply and demand can determine an equilibrium price. Protection of both duckweed and fish producers' interests, therefore, assumes some formal link- age between duckweed and fish production. Figure 23 illustrates product flows and linkages in a model of duckweed production and utilization. Several simple models are discussed below. Demand models The simplest duckweed/fish production par- adigm is a demand model, in which the fish producer expresses de- mand for duckweed with an offer to pay a floor price for all the supply brought to him. This mechanism was tried in Khulna, Bang- ladesh, to foster the collection of naturally occurring duckweed from village ponds. It had the effect of stimulating deliveries of fresh duckweed by villagers while wild stocks lasted. But, having deplet- ed existing duckweed stocks, villagers did not, as expected, request technical assistance to develop and maintain duckweed culture ponds. Supplies of duckweed quickly dropped to levels insufficient to maintain a duckweed-fed carp fishery. and an increase in the of- fering price had little effect on supply. A more active model in which duckweed farmers are provided with technical assistance and investment capital, in addition to a floor price offer, is likely to produce better results. Without guaran- tees on either side, however, duckweed producers retain little pric- ing leverage and remain vulnerable to arbitrary termination, while fish farmers are vulnerable to supply uncertainties. Two-unit linkage Paired linkage between a duckweed farmer and fish farmer, reinforced by formal short-term agreements spec- ifying mutual obligations with respect to price and supply, is more satisfactory, both from a productivity as well as an equity point of view. By enabling formal negotiation, this mechanism allows better distribution of benefits between the two parties. However, simple linked production may not provide an adequate buffer against fluc- tuations in duckweed supply and demand. Group linkage Close linkage between and among two producer groups appears to provide the best circumstance for duckweed/fish 42 Inputs: Fertilizer, Water, Wastewater Duckweed Farm Duckwed Fish Farm Ding Duckweed Meal Ponds & Cages Mixed Feed Additives: Maize, Preparation Wheat & Vitamin Premix Fish Harvesting P ial & Processing Feed Feed Pondside Bulk Sale of Fish Sales LMeat Eggs Pelleted Feed Eggs Sale in Village Urban & Small Town Commerclal Markets Figure 23. Product flows in integrated farming of duckweed, fish and poultry production. Pooling of supply and demand is the major difference be- tween this and the paired producer model. The supply buffer can also be augmented in a group context by guaranteeing adequate substitution, for example, water hyacinth, in the event duckweed production does not meet some specified minimum. Fish producers 43 should also provide guarantees for floor price and minimum quanti- ty purchases. The possibility of substitution within each group- duckweed production for fish production and vice versa-provides dynamic tension to price negotiations and therefore higher returns to duckweed producers. Vertical integration Vertical integration is a logical response to the uncertain relationship between duckweed producers and fish producers, but it is unclear at this point whether farmers will prefer separate or integrated operations. Because duckweed production has significantly lower net returns than does fish production under existing pricing arrangements in Banglasesh, fish farmers who could vertically integrate may find it more attractive to devote all their production capacity to fish farming while working to stimulate production of their duckweed requirements among neighboring farmers. Entry into duckweed production by fish farmners may also be inhibited by the need to hire labor and somewhat lower produc- tivity compared with an owner-operated duckweed farm. Poorly paid hired laborers are unlikely to sustain either the level of effort, or to develop the sensitivity to crop fluctuations that are essential to maintain high duckweed productivity. For most duckweed farmers, moving to a vertically integrated production model is unlikely because of significantly increased risk and the requirement to defer gratification. To achieve such produc- tion integration, duckweed producers must also gain access to ad- equate land area, infrastructure, and working capital to sustain at least six months of production. It is likely that they would also have to forgo the daily salary-like cash reinforcement derived from duck- weed production contracted to a nearby fish farming operation. Linkage catalysts Duckweed production is technically com- plex and there are large requirements for working capital for joint duckweed and fish production. It is critically important to coordi- nate between both production elements, yet there are few operating production centers that can serve as models for aspiring produc- ers. For these reasons, it is important to develop an effective institutional framework for stimulating and managing duckweed and duckweed-fed fish production. It is unlikely that farmers or groups of farmers will band togeth- er of their own volition in a coordinated duckweed/fish venture. An 44 external catalytic agent is required. This can take many institution- al forms: government extension services, private voluntary agen- cies, producer cooperatives, or agribusiness. The agency's primary responsibility should be to ensure smooth coordination between duckweed and fish producers. Also, the agency should ensure that adequate supplies of working capital and technical assistance are available. Efficient duckweed production requires continuous su- pervisory, technical, and financial reinforcement. Technical assistance and extension issues Unlike tradition- al crops which need only sporadic attention, duckweed cultivation, and duckweed-fed fish culture are both continuous production processes. Duckweed production, in particular, departs signifi- cantly from the conventional agricultural paradigm of planting -> fertilizing/crop maintenance -> harvesting -> processing -> storage - > sale spread over a growing season ranging from two months to two years. All of these elements are compressed into a daily cycle in duckweed farming. Adapting to farming as a continuous process is likely to demand a difficult conceptual adjustment on the part of most farmers. Receiving daily payment for daily production is strong rein- forcement for good practice. A farmer who fails to fertilize, main- tain, or harvest his crop adequately will experience an immediate drop in production and, consequently, in income. He would not have to wait for three months before facing the consequences of his action. Feedback is immediate and has a salutary reinforcing effect on both quality and level of effort. The duckweed-fed fish culture model discussed here has also been structured as a continuous production process. Feeding with duckweed is continuous throughout the day, while guarding and monitoring of the fish crop continues both day and night. Only har- vesting is conducted periodically. The role of a village level extension agent is to ensure: (1) that each participating farmer is trained in the latest duckweed or fish farming techniques; (2) that he understands the continuous nature of the production processes; (3) that he continues to engage in good practice; and (4) that he continues to receive immediate payment for his daily product. This suggests that extension support for duckweed and duckweed-fed fish production should also, as with 45 the processes being supported, become a continuous process. And it suggests that duckweed extension should have some financial and commodity exchange capability. Building these elements into existing extension systems, whether government or private, is likely to be difficult. Extension, credit, and commodity exchange capabilities are more appropriate- ly built into duckweed research and demonstration centers. These centers of applied research could then evolve into integrated cent- ers for duckweed research and dissemination. Credit requirements Credit support for duckweed and duck- weed-fed fish farming is essential. Both are intensive processes that need a steady flow of investment. Credit for these linked proc- esses is characterized by two features: (1) it is appropriately dis- bursed continuously in small, productivity-based increments, and (2) it is considerably greater than the credit required for compara- ble conventional farming processes. Where wastewater is the source of water and growth nutrients for duckweed production, lower recurrent costs mean that credit requirements will be about half those for hydroponic culture of duckweed. The performance of agricultural credit programs to small farm- ers worldwide is poor. Loan amounts seldom match real require- ments: disbursements are slow; recovery rates are low; and where recovery is mandatory, as in the United States, small-farmer bank- ruptcies are commonplace. While a discussion on the reasons for this poor performance is beyond the scope of this paper, common belief holds that, beyond the more frequently cited structural defi- ciencies of the credit institutions themselves, a primary failing of ag- ricultural credit programs is the inability of farmers to manage their credit. Experience shows that farmers are likely to consume directly a significant portion of the credit they receive, and the greater the amount of disbursement, the higher the proportion of consumption. The amount of working capital required for linked duckweed and fish production is high by most comparable standards. Ferti- lizer inputs to duckweed production are higher than for other crops. Similarly, duckweed input requirements to a carp fishery are higher than those of comparable fish feeds. Both require daily in- puts and, therefore, a continuous flow of cash. Assuming that the duckweed farmer will be paid immediately for his daily product, 46 credit requirements for working capital may then be focused direct- ly on the fish farmer. At the beginning of his production cycle, he must have access to sufficient working capital to enable daily pro- curement of duckweed supplies for six to seven months. At a price of $0.03/kg for fresh duckweed, a farner growing one hectare of carp will require between $1,500 and $1,600 for a year's supply of duckweed. This is for most Bangladeshi farmers more than their expected household income for a year. The likelihood of their re- taining the money over six months and spending it for duckweed procurement is, therefore, very low, and a phased supply of incre- mental installments is needed. In the case of duckweed/fish production the risk for farmers can be reduced through close technical and managerial involve- ment by the credit institution. A village-based agent should man- age the exchange of duckweed between duckweed farmers and fish producers, and should also arrange direct payment to duckweed producers on behalf of fish farmers. Direct payments to fish farm- ers should be for (1) salaries of external labor employed directly in fish production, and (2) sustenance allowances to fishery owners. Credit institutions should serve as exchange agents in the final disposition of fish. Income from fish sales should flow through the credit institution before net payments are made to fish producers. In performing these exchange services credit institutions should add value to the production processes by improving both production and marketing efficiencies, and by continuously reinforcing good practice through technical assistance and efficient timing of f'miancial inputs. Pricing issues Current experience suggests that at a price of 1.0 Taka. or about $0.03 per kg, a duckweed farmer in Bangladesh can expect to net less than one-third of what a fish farmer can earn from the same amount of land. Close linkage of duckweed and fish production is likely to place continued upward pressure on the price of fresh duckweed. This upward pressure is moderated by a general acceptance that fish farmers deserve a higher return be- cause they accept greater risk and make higher capital invest- ments. Upward pressure on the price of duckweed is also relieved slightly by the threat that fish farmers might decide to vertically in- tegrate their operations by producing duckweed themselves. 47 Where extension-credit institutions provide linkage services be- tween duckweed and fish producers, provision should be made for a mechanism to negotiate the price of duckweed when fluctuations are justified to distribute profits from linked production more equitably. As a market for dried duckweed meal is gradually established, pricing of fresh duckweed will be influenced more by market prices of dried duckweed meal and protein extract. And these will, in turn. be tied closely to prices of competitive products derived from soy- bean and fish. Profitability The projected rates of return on investments in duckweed-fed carp production and duckweed production compare favorably with alternative investments in the agricultural sector in Bangladesh. Annexes 1 and 2 estimate the profitability in Bang- ladesh of five-year investments in duckweed-fed carp culture and duckweed production respectively. The profitability of duckweed production is especially sensitive to two factors, (1) the cost of fertilizer, and (2) the sale price of fresh duckweed. Where all fertilizer and most water are obtained from a domestic wastewater stream, the internal rate of return on duck- weed production escalates from 44 percent to 63 percent. A 30 per- cent increase in the price of fresh duckweed brings the internal rate of return up to 74 percent. The profitability of duckweed-fed fish production is most sensitive to the price of fresh fish, and the cost of investment cap- ital, but reasonably insensitive to the price of fresh duckweed. A 30 percent decrease in the price of fish reduces the internal rate of return to 45 percent. However, a 30 percent increase in the price of duckweed only reduces the internal rate of return from 85 percent to 80 percent. 48 Section 5 - Duckweed-Based Wastewater Treatment Systems Effective treatment of nightsoil and wastewater, at both the vil- lage and urban level, remains an elusive objective in most develop- ing countries. While there are many reasons for this, experts generally agree that the overriding factor is cost. Conventional treatment systems, which generally rely on heavy aeration are pro- hibitively expensive to install, and both difficult and costly to oper- ate and maintain. If affluent cities such as Sydney and San Diego cannot afford to make the billion dollar investments required to provide effective treatment of their wastewater, what prospects are there for Calcutta or Lima? And if Lima and Calcutta, which have viable municipal governments, cannot afford wastewater treatment how can it be accomplished in small towns and villages which have essentially no tax base? Duckweed-based wastewater treatment systems provide genu- ine solutions to these problems. They are inexpensive to install as well as to operate and maintain. They do not require imported com- ponents. They are functionally simple, yet robust in operation; and they can provide tertiary treatment performance equal or superior to conventional wastewater treatment systems now recommended for large-scale applications (for terminology see Box 1). Finally, and perhaps most importantly, duckweed wastewater treatment sys- tems have the potential, by turning wastewater into valuable duck- weed meal, to return a net profit against both capital and recurrent costs. This being the case, it suggests that in future, cities like Lima and Calcutta cannot afford not to treat their wastewater. Duckweed wastewater treatment systems remove, by bioaccu- mulation, as much as 99 percent of the nutrients and dissolved sol- ids contained in wastewater. Duckweed systems distinguish themselves from other efficient wastewater treatment mechanisms in that they also produce a valuable, protein-rich biomass as a by- product. Providing accumulated toxin and heavy metal levels are not high,' the harvested duckweed biomass may be used as the 'An unlikely occurrence in village, small town and urban periphery systems. 49 Box 1. Wastewater Treatment Origin of wastewater In cities and urbanised areas sewers and, sometimes, open-air drains con- vey wastewater from the households, from workshops or even some factories. The sewers often also drain away rain water to prevent flooding. This municipal sewage has a more or less typical composition all over the world, though locally characteristics reflect the activities in the drainage area. Parts of cities which are not connected to the sewer network, and towns and villages without sewers. have to rely on open channels to drain kitchen sullage, and on septic tanks and soaking pits to have the heavily contaminated toilet ('black') wastewater percolate into the soil. Factory discharges are very specific per industry sector with regard to their quantity and composition. Wastewater composition and purpose of treatment The following categories of water contaminants can be distinguished: - organic substances which can be degraded by bacteria with the help of oxygen dissolved in the water (BOD); too much of these substances de- pletes oxygen in the water rendering it septic and unfit for human or an- imal use: - nutrients (nitrogen and phosphorus) which make plants and algae grow profusely to the extent they endanger normal use of the water (eu- trophication): - heavy metals and organic micropollutants, generally from factory dis- charges, which may be toxic ('toxins') to plants, animals and humans: and - pathogenic (disease causing) micro-organisms which abound in hu- man excremental and black wastewater. The purpose of treatment can be public protection (keeping pathogens away from the habitat, or killing them by disinfection). protection of en- vironmental quality (removing oxygen consuming substances, and tox- ins) or achievement of water quality and ecological standards (nutrient removal). Elements of uNastewater treatment 'Primary' treatment aims at removing settleable and floating matter in a simple settling tank: up to 40 percent of pollution can be removed (but not all BOD nor nutrients or toxinsl. 'Secondary' treatment conventionally uses bacteria and forced oxygen adductioni to remove the remainder of oxygen consuming sub- stances and is traditionnally the most expensive component of the system. Only the additional 'tertiary' stage. applying chemical or biological methods, is also able. at a considerable expense, to remove nutrients and part of the toxins. Final- ly, pathogen reduction can only be achieved fully by 20 - 25 days impoundment. as in a duckweed system, or by chemical disinfection. 50 sole feed input for fresh-water pisciculture, and as up to 40 percent of poultry feed. This biomass might also be useful for a variety of other domestic animals. Duckweed wastewater treatment systems are, at their core, la- goon systems. They differ from conventional lagoon systems, how- ever, in that they (a) achieve a significantly higher level nutrients removal from the wastewater stream; and (b) achieve removal of ox- ygen consuming substances and pathogenic organisms to an ex- tent comparable to algae based lagoons; but (c) without having the disadvantage of large amounts of algae being washed out of the sys- tem as suspended solids. The effect is to produce a high-quality effluent which can halt or significantly reduce the continual influx of harmful substances (nitrogen, phosphorus, etc.) into receiving bodies of water (rivers, lakes or seas). Unlike conventional lagoon systems, duckweed wastewater treatment systems also have a low algal content - there- by meeting the most stringent discharge requirements for suspend- ed solids. Duckweed system discharge contains few organic compounds and may therefore be chlorinated without significant production of carcinogenic trihalomethane compounds. Finally, because they are more efficient than conventional lagoon systems duckweed systems occupy less (expensive) land to achieve a higher level of treatment.2 The section which follows describes the pilot duckweed wastewa- ter treatment plant at the Mirzapur experimental site which has been in operation since July 1990. Results have been impressive, though the system has not been optimized-notably the required surface may prove lower than the present situation. Treating an average flow of 125 2Duckweed systems can vary in depth from half a meter to three or more meters, but this requires further study. System design should balance detention time-for pathogen reduction-against land and excavation costs, nutrient removal targets and the feasibility of chlorinating the final effluent. All things being equal a duck- weed system having the same depth and detention time as a conventional faculta- tive lagoon system will provide a higher overall level of treatment-for pathogen, nutrient and suspended solids removal. Secondary effluent standards can be achieved with between 0.3 and 1.0 m2 per person, and advanced tertiary standards with 2.5 to 4.0 m2 per person. This contrasts well with the facultative lagoon requirements of between 1.8 and 2.8 m2 per person to achieve close to secondary e!lVuent standards. 51 Table 2. Quality of final treated effluent for March 23, 1991 Mirzapur Experimental Site BOD5 NH3 P 7urbidity Treatment Phase (rng/l) (mg/i) (mg/I) F7TJ1 Raw influent 120 39.40 1.90 113 Primary 60 32.20 2.00 85 Duckweed 1 0.03 0.03 10 US Summer Standards: Washington D.C. area 10 2.00 1.00 202 1. This turbidity unit standard is roughly equivalent to total suspended solids (TSS) times two. 2. Standards for the Patuxent Valley in Maryland, north of Washington D.C. TSS standards of 10 mg/l are shown in FTU units (20) for comparability. m3/day of hospital, school, and residential wastewater produced by a population of between 2,000 and 3,000 persons, the 0.6 hectare plant produces a final treated effluent which exceeds the highest quality standards mandated in the United States.3 Table 2 shows typical influent, primary effluent and duckweed system effluent data for the Mirzapur experimental wastewater treatment plant. Many other wastewater treatment facilities designed solely for municipal treatment in the U.S. and elsewhere have produced bet- ter than secondary effluent quality for flows ranging from a few hundred cubic meters per day to over 30,000 m3/day.4 Even higher flow rates are being designed for large cities with hundreds of thousands of inhabitants. These systems have been designed to conform to all standards of design and operation im- posed by the U.S. Environmental Protection Agency and other sim- ilar regulatory agencies in various countries. 3 The wastewater effluent from the Kumudini Hospital complex (Mirzapur), with BOD of 120 mg/I, is not typical of most developing country wastewater streams which are commonly more polluted. The collection system at the complex does not capture a significant portion of the discharge from the complex-particularly kitchen wastes, and hostel septage-and the water discharge from the hospital itself Is significantly higher than average institutional dtscharge in developing countries. This contributes to both a low flow and a relatively low BOD. 4 The Lemna Corporation. St. Paul. Minnesota. U.S.A. 52 The basic mechanism employed by the duckweed wastewater treatment system is to farm various duckweed species on the wastewater requiring treatment. The rapidly growing plants act as a nutrient sink, absorbing primarily nitrogen, phosphorus, calci- um, sodium, potassium, magnesium, carbon and chloride from the wastewater. These ions are then removed permanently from the ef- fluent stream as the plants are harvested. Depletion of nutrients causes diminished duckweed growth. The starved plants then begin processing increasingly greater amounts of water as they search for growth nutrients. In the proc- ess, they absorb virtually every chemical present in the wastewater stream. The small volume of plants harvested during this polishing process may contain unacceptably high levels of toxins and heavy metals when influent contains a significant volume of factory dis- charge. If so, they should be disposed of as green manure for crops and not used as food or forage. In such situations the duckweed system should be operated to optimize the combined "value" of achieved effluent quality and duckweed crop. Maintenance of efficient duckweed growth requires even distri- bution of a thick layer of plants across the entire lagoon surface. This has the additional effect of shading the water below from sun- light and preventing growth of algae.5 Harvested duckweed plants contain up to 45 percent protein by dry weight and may be used without further processing (i.e., drying) as a complete feed for fish. Dried duckweed meal can pro- vide the protein constituent of various mixed animal feeds. The vi- tamin A and pigment content of duckweed have proven particularly valuable in poultry diets.6 A typical duckweed wastewater treatment plant will yield a dai- ly harvest of up to one ton of duckweed plants (wet weight) per hec- tare. This can, in turn, produce up to 100 kilograms of fish or 90 kilograms of dried, high-protein duckweed meal each day. 5 Algae are the major constituent of TSS in the final effluent of most wastewater treatment systems. 6 This has been shown during 4 years of research on the nutritional value of Lem- naceae conducted by The PRISM Group in collaboration with the Agricultural Uni- versity of Peru and the Ralston Purina company. See Haustein et al. 53 Primary system The primary phase of the duckweed wastewa- ter treatment system in a simple and cheap basin, which receives all the raw wastewater influent. This phase in itself is a quite com- mon primary sedimentation, but should be designed to release the maximum amount of nutrients from the settled matter: in the sub- sequent phase, duckweed will thrive on these nutrients. Like any primary treatment process, the principal objective is to separate floating material and achieve significant solids removal through sedimentation-all at a low capital cost. Sedimentation Achieving efficient sedimentation is important to prevent degradation of initial duckweed treatment runways. Septage and influent wastewater must also be introduced with minimal aeration to maintain anaerobic conditions and to avoid odor nuisance. This is easily achieved using a deep tank or pit, and is enhanced by maintaining methane storage under slight pres- sure. A deep, reinforced circular tank with a vertical, centrally lo- cated, low-pressure, large diameter inflow pipe will achieve efficient settling while also maintaining anaerobic conditions within the tank.7 Twin primary tanks are usually necessary. Both tanks should be located side-by-side with the first tank being built approximately 30 cms above the second tank to enable gravity flow-through. Ini- tially, both systems should be operated in series, with the second tank receiving the effluent from the first tank. As sedimentation in- creases in the first tank, efficiency will also drop and an increasing volume of sediment will be passed through, and be trapped by, the second tank. When total fluid volume in the first tank has been re- duced by 50 percent it should be bypassed, with all influent flows passing only through the second primary tank. The first tank should then be drained and sludge removed by whatever mecha- nism is most safe and efficient given local circumstances. The cleaned tank (#1) should be brought back into service as soon as possible. Eventually, the second tank (#2) will also require 7 Where cost factors prevent construction of deep. enclosed sedimentation/diges- tion chambers and odor control is not considered to be a high priority, reasonably efficient sedimentation can still be achieved using two deep, open earthen tanks. Inflow should be designed to minimize turbulence and aeration. 54 desludging. This will, of course, require temporary bypassing the tank, with direct discharge of primary effluent from tank #1 into the duckweed plug-flow system. Sludge disposal Sludge should be analyzed for toxin and heavy metal concentrations prior to project implementation. If found to meet established criteria, the project should include a mechanism for composting sludge and either using it directly or selling it as garden manure. Sludge should otherwise be disposed of in a manner which will minimize entry of toxins or heavy metals into the human food chain. The most profitable application is likely to be use as a fertilizer in a nearby agroforestry project. Primary treatment must deal with two types of floating materi- al, plastic and flotsam carried on the raw influent, and scum-like material floated from the bottom in anaerobic systems. Flotsam is easily removed through course screens. Scum is trapped by releas- ing effluent from the primary tanks 0.5 meters below the surface. The resulting crust of floating material will also serve to minimize surface aeration and reduce odor in open-cut primary systems. Odor control Both primary settling tanks should be covered if possible. The resulting odor reduction will have a significant sal- utary effect on acceptance of the facility by persons having occasion to live or work near the facility. Efficient operation of the primary facility dictates maintenance of an anaerobic system. As such, generation of a significant volume of methane and hydrogen sulfide is inevitable. These gases should be trapped under an airtight cover and either used as biogas or flared-off. Bad smells are the most frequently cited objections of people living in the vicinity of wastewater treatment facilities. Designed and operated correctly, a duckweed primary system should issue no objectionable odors. In fact, a well landscaped duckweed waste- water treatment makes an excellent park. The Mirzapur facility is favored by local couples as a meeting place. Costs System cost is an important consideration in the design of a primary process for a duckweed wastewater treatment system. Should cost prove to be a significant constraint it is possible to achieve effective primary treatment with two simple open-cut fac- 55 ultative lagoons. Unlike the closed system described above, open systems may present significant public relations problems to the operating agency. In villages and rural towns without sewer infra- structure, the primary phase in the duckweed system may be de- leted. A structure should be organized which motivates village dwellers to make use of well designed and maintained latrines sit- uated on the banks of the duckweed lagoon. The purpose is to get as much excreta, containing valuable nutrients directly into the la- goon rather than having them deposited in the neighbourhood or in pit latrines. The system consists of just one deep pond for duck- weed production; excrements settle down quickly to the bottom where they gradually decompose. Of course, crop collection re- quires more careful procedures to prevent contamination. Duckweed plug flow system The essential element of a duck- weed wastewater treatment facility is the duckweed system itself. It consists of a shallow, lined8 pond system designed to allow effective cultivation of duckweed plants and incremental treatment of a wastewater stream. As such, the system must enable efficient har- vesting and maintenance of the duckweed crop while also prevent- ing short-circuiting of the wastewater flow. The duckweed plug flow system may be thought of as contain- ing two distinct elements: (a) the duckweed farm; and (b) the waste- water polishing facility. Under circumstances where wastewater consists primarily of domestic sewage, these two elements may be indistinguishable. Duck1weedfarin The principal objective of the duckweed farm is to produce as much usable, harvested duckweed as possible while also maximizing net returns from the process. In so doing, the objective of achieving maximum removal of nutrients from the wastewater stream is also achieved. Like all biological systems, duckweed plants prefer certain growth conditions over others. Maintenance of these conditions is important in achieving both efficient plant growth and effective wastewater treatment. 8 Lining is essential to both prevent water loss and protect aquifers. Unlike the multilayer linings strictly mandated for landfill sites in North America and Europe. a relatively inexpensive clay lining will usually suffice. 56 While duckweed species are known to survive under widely varying conditions of both water temperature and chemistry, their rate of growth is quite sensitive to variations of both.9 Recirculating systems The ultimate treatment objective of re- moving all nutrients from wastewater inevitably leads to duckweed starvation at some stage in the treatment process. This eventually leads to virtual cessation of plant growth. At the other extreme, high loadings of nutrients (ammonia in particular), surfactants1o and compounds with herbicidal properties can have a similar effect but this is easily prevented. This is achieved by recirculating a por- tion of the final treated effluent. Systems should therefore be de- signed to begin and end at a proximate location. This makes recirculation a simple matter of lifting treated effluent about six inches and pumping it a short distance. A simple rule of thumb for dilution of primary effluent is to ensure that BOD5 at the head of the first duckweed treatment runway is maintained under 80 mg/I. The objective of maximizing minimum surface temperatures and minimizing maximum surface temperatures is served by in- creasing system depth and stimulating system mixing. An additional consideration dictating system depth is total detention time (approx- imately 20-30 days to achieve acceptable pathogen reduction). Experience suggests that systems with a maximum operation- al depth of 1.0 meter can provide acceptable temperature buffering and detention time without incurring unacceptably high costs. I 1 Distributing and containing duckweed plants Among fac- tors affecting duckweed growth, unconstrained access to the pond surface ranks as the most important. Plants should be distributed across the entire surface to make full use of the productive potential 9 Refer to Sections one and two for specific information on optimal conditions for duckweed cultivation. 10 Surfactants are a product of soap and detergent in effluent streams. In high concentrations they can "dissolve' duckweeds' protective waxy coating, leaving plants more vulnerable to fungal infection. 1 l Depths of between one half and three meters are also acceptable. For example, a circumstance with relatively low BOD and high land costs and a requirement to maximize pathogen removal would be designed with deep runways and low reclrcu- lation. Similarly, a situaUon with high BOD and inexpensive land might be better served by an extensive, relatively shallow system with high rates of recirculation. 57 of that surface. They should also be distributed in a manner which does not constrain their growth. Increasing the base population of plants in a given area increases the multiplicative potential of that population. There is, however, a point of diminishing returns, where the inhibitive effect of crowding on plant reproduction outweighs the increased productive potential of a higher base population. Efficient distribution of duckweed plants across the entire available growing surface requires that plants be contained in rel- atively small, discrete cells. This is achieved by two means: (a) plac- ing an interlocking floating grid over the ponds or runways used for growing duckweed; or (b) building containment cells with low earthen berms and bunds. Choice of containment system is prima- rily a function of land, labor and material costs but is also influ- enced by factors such as prior circumstance12 and choice of system operation intensity. Floating containment structures should be lWV resistant and sufficiently robust to survive 5 or more years of heavy harvesting activity by self-propelled mechanical harvesters. Where capital is constrained containment booms can be fashioned from large diam- eter bamboo or some other inexpensive floating material. They will, however, require frequent replacement, and will probably cost more in the long run than barriers made from extruded plastic. The size of the grid is determined by mean ambient wind conditions and the maximum projected system flow velocity. Cell sizes on existing PRISM and Lemna Corporation wastewater treatment systems range between 25 m2 to 50 M2. Alternatively, low earthen berms are also effective in creating efficient duckweed production cells. This system, depicted in Fig- ure 25, allows use of perimeter harvesting with a variety of hand tools and small mechanized harvesters. Berm systems have the ad- ditional advantage of providing increased area for collateral crops which can significantly boost total system profitability. Harewsting Having determined the standing crop density which realizes the highest duckweed productivity, efficient management dictates maintenance of a steady state system at that density. Each 12A duckweed system 'retrofit" on an existing lagoon will typically use an extensive floating grid for containment. 58 DUCKWEED CULIVATION CHANNELS - 20 melers X 5 molers X 1.0 meers EFFLUENT ReeircLking _ . i .1 ..... em ?NFLUENT [Harvesling ber s ilh Emergenl Growlh - 2 melers ide| OPERATING PARAMETERS T -' 1. SYSTEM CAPACITY: 50, 000 liters per day ot 150 mg/I BOD influent 2. RECIRCULATION CAPACITY: 2 times totol system flow 3. TOTAL SALTS REMOVAL: 1, 500 grams per day (at 22 degrees Centigrade +) WATER QUALITY MONITORING SITES SLUDGE DRYNG AND COWPOSING BEDS FLOATING CONTAINMENT BOOM The PRISM Croup Figure 24 Model duckweed wastewater treatment system using floating containment barriers cell should be harvested back to the target density. Optimal system densities on existing duckweed systems range from 400 to 800 grams of duckweed per square meter. The choice of harvesting technique is dictated by system con- figuration as well as the cost of labor and capital. The most simple harvesting mechanism involves scooping of plants from the pond surface using handtools. This mechanism is facilitated by a facility design which enables harvesting from a perimeter surface-typical- ly a narrow plug flow system. Larger, broader pond-based systems require harvesting from self-propelled craft. These may be either engine driven, or powered by the harvesters themselves. In most developing country applications, systems should be designed to en- able labor-intensive perimeter harvesting. Regular harvesting is important not only because it generates a valuable biomass byproduct, but also because bioaccumulation remains the principal mechanism of wastewater treatment, and harvesting ensures that the accumulated nutrients and toxins are 59 EFFLLENT Diniding Berms with Emergent Growlh NFLLIENT |RecirCUbtliDn >[I Sm x4m x 0.5m' l , ~~~~Pump ...WW Ef Duckweed Pfoduction _ Recirculationl .IX m mllll,11111 rColl '5m x 5m| | t ,syslern Moli\ Urn Dividing ond I~~~~~~~~~~~M . .... .... . 4##H Haryesting Brerm wifh Permcnent Emergent Growth el'imng (I [ X 2m x t5m] | ,,cnks ....G ;t ...Iae-j IConnecting Pipe Maitn DAuckeed tar vestingr Lone Primrry Bypass Prlmary Tanks Ouckweed Production Celos [0.4m lo Oinm deep] [L x I ml [for desludgmng| [5 to 6 hours detention] - - -- Skuge Drying ond Conposting 3ed& Th P°ISM Group Figure 25 Model duckweed wastewater treatment system using earthen berms for crop containment permanently removed from the wastewater being treated. Harvest- ing is also important to maintain a healthy, productive crop. Younger plants not only maintain a better nutrient profile (i.e., they contain more protein and less fiber), but they also reproduce and grow more quickly than older plants. Algae shade A significant benefit of duckweed systems over other lagoon-based wastewater treatment systems is that they are capable of efficient removal of influent suspended solids and they prevent formation of algal suspended solids which are the bane of lagoon system effluent. This is achieved through the simple mech- anism of shading. A dense layer of floating duckweed plants pre- vents sunlight from reaching algae populations distributed throughout the water column. Unable to photosynthesize they sim- ply die and precipitate to the pond bottom. Systems with enclosed primary treatment units maintain algae inhibition from the outset and will provide marginally better total suspended solids (TSS) removal than systems with open primary lagoons. In either case, duckweed wastewater treatment systems can consistently bring total final effluent TSS to below 5 mg/liter. 60 Nutrient uptake efficiency Duckweed plants are remarkably efficient at removing elements which are, for them, growth nutri- ents. These include some organic compounds, as well as ions of el- ements such as nitrogen, phosphorus, potassium, magnesium, calcium, sodium, chlorine, boron, and iron, among others. Duck- weed can directly remove both complex carbohydrates such as su- crose, fructose and glucose, as well as organic nitrogenous compounds such as urea and most amino acids. While growth nu- trients remain, duckweed plants are disposed to absorb them to the exclusion of other elements present in the wastewater column. As such, well-fed duckweed plants cannot be considered ideal en- gines for complete treatment of wastewater13 with high toxin and heavy metal content. Sqfety of har-vested duckweed plants Duckweeds' predilec- tion for exclusive uptake of nutrients is important in enabling the safe utilization of plants harvested from urban wastewater. Testing, over the years, of many duckweed plant samples harvested from nutrient rich urban wastewater has consistently failed to find any heavy metals or known toxins in concentrations approaching USF- DA (United States Food and Drug Agency) food standards prohibit- ing human consumption. 14 Duckweed wastewater polHshing Effluent from a polishing treatment plant has normally received a series of expensive treat- ments. Duckweed plants do, also, provide a comiplete wastewater treatment engine. Starved duckweed plants-i.e., plants unable to find sufficient nutrients to maintain rapid growth -undergo a re- markable metamorphosis: plant protein drops below 20 percent; fiber content goes up; roots become long and stringy; fronds be- come larger and discolored; and, most importantly, the plants be- gin processing huge amounts of water in their search for sustaining nutrients. In the process, they absorb virtually everything still present in the wastewater. 13 Duckweed wastewater treatment systems are, nevertheless, capable of efficient toxin and heavy metals removal in a polishing process described under 'Duckweed Wastewater Polishing" below. 14 Haustein, A.. R. Gilman, P. Skillicorn, 1987, The Safety and Efficacy of Sewage- grown Duckweed as feed for Layers, Broilers and Chicks. report to USAID Science Advisor. 61 Polishing units, necessarily, form the final stage of a duckweed wastewater treatment system. The polishing function takes place in the final stretch of a duckweed wastewater plug flow system. In in- stances of wastewater with heavy concentrations of toxins and heavy metals, the beginning of the polishing zone should be explic- itly indicated. Plants harvested from the zone should then be dis- posed of in an appropriate manner. Most wastewater does not, however, contain significant concentrations of either toxins or heavy metals, and polishing zones may simply be considered to be the latter reaches of a continuous duckweed treatment process. Harvest volume and plant quality will be somewhat lower than that achieved from the bulk of the farming zone, but polishing plants need not be excluded from the main harvest. Pathogen removal Pathogen reduction in any lagoon system relies on two simple mechanisms: sedimentation and die-off. Para- sites and parasite ova precipitate with other suspended solids and are trapped in the bottom sediment. Other pathogens, suspended in water, simply die as a function of time and temperature. A suffi- cient detention time must be provided to ensure die-off of patho- gens adequate to meet effluent discharge or reuse standards. As with any organic surface area enhancing material intro- duced into wastewater, duckweed plants do marginally concen- trate pathogens on their surfaces. As such, pathogens will, inevitably, be harvested along with the duckweed crop. If harvest- ed plants are used green as fish feed, these bacteria experience even greater dilution and faster die-off in the fish pond. The small number of surviving pathogens consumed by fish will be digested in their guts. In instances where plants are processed and dried, desiccation will achieve even more rapid die-off. No viable human pathogens could be cultured from dried sewage-grown duckweed meal in 4 years of testing. 15 15 Haustein, A., R. Gilman, P. Skillicorn, 1987, The Safety and Efficacy of Sewage- grown Duckweed as feed for Layers, Broilers and Chicks. report to USAID Science Advisor. This research, conducted in collaboration with enteric disease experts from The Johns Hopkins University, examined both wet and dried Lemnaceae har- vested from the San Juan wastewater lagoons located in Lima, Peru, for presence of various human enteric pathogens. 62 A secondary advantage of the duckweed system in this respect lies in the very low concentrations of suspended and dissolved or- ganic matter in its effluent, when compared to regular algae based treatment lagoons. As described earlier, removal of organic pollu- tion takes place efficiently. In addition, growth of algae, always hard to remove from water, is inhibited by the shade created by the duckweed layer on the pond surface. If the necessity arises to pro- duce an effluent totally free of pathogens, such effluent can be dis- infected safely by chemical chlorination; chlorination of water containing too much organic substances produces carcinogenic tri- halomethane, which should be avoided. Final effluent discharge Under most circumstances the final effluent from duckweed wastewater treatment systems will be su- perior to the receiving stream or waterbody. Duckweed system run- off may therefore be used as input to virtually any water-intensive operation-irrigation, factory use and cooling systems, among oth- ers. Providing thorough filtration 16 and some form of disinfection is performed-either chlorination, ozone or ultraviolet treatment- treated effluent from a duckweed system may potentially be used as input to municipal water supply systems. In water constrained areas such as the Middle East, the Caribbean and the west coast of South America, this represents a viable, ecologically superior alter- native to desalination and costly dam and aqueduct projects. Commercial systems In the United States, a commercial duckweed based wastewater treatment process has been approved by the U.S. Environmental Protection Agency for funding in munic- ipal applications, which has now been applied, under varying con- ditions, in over sixty distinct locations throughout the United States, Europe and Latin America. The treatment system consists of a sophisticated interlocking network of floating booms and hy- draulically-driven mechanical harvesters to enable the growth and harvesting of duckweed on vast open ponds. These treatment facil- ities routinely achieve secondary to tertiary effluent standards for municipal waste streams in climates varying for sub-arctic to trop- 16 Simple slow sand filters have been shown to provide excellent removal of organic compounds and are now routinely recommended as pretreatment for water treat- ment plants that draw from surface water sources. 63 ical. Such systems compete favorably against mechanical wastewa- ter treatment systems'7 on both capital costs and treatment efficiency. In addition, the operating requirements are much less demanding than those of conventional systems, resulting in sub- stantially lower energy and labor costs. In general, care must be taken to ensure that the design, con- struction and operation of any wastewater treatment system con- forms to the local regulations and design standards. This ensures protection of public health, public safety and the environment. To this end, it is advisable to retain the services of professionals in the wastewater treatment field to advise and assist in such design and construction programs. 17 Aerated lagoons, activated sludge, or high rate algae systems. 64 Section 6 - Alternative Uses for Duckweed, Constraints and Future Research Developing alternative uses for duckweed Use of duckweed is currently restricted to processes that can utilize freshly harvest- ed plants. Further, transportation and storage constraints dictate that these processes be near the duckweed farm. Nutritionally, dried duckweed is an excellent substitute for soybean meal and fish meal in a variety of products. However, the economic potential of the duckweeds may not be fully realized until they can be eco- nomically reduced to a dried, compact commodity. This requires drying, and either pelleting, or powdering. All drying technologies consume large amounts of energy, which is expensive, except waste heat and solar energy. Desiccat- ing duckweed, which may contain from 92 to 94 percent moisture, using purchased energy - either gas, oil, electricity or biomass- is not economically feasible. If duckweed is to become a traded com- modity, drying must be achieved through efficient application of ei- ther solar or waste process heat. Duckweed plants have a waxy coating on their upper surface that is a good binding agent for pelleting. Dried meal, fed through conventional pelleting equipment, either alone or in combination with other feed ingredients, produces an excellent pellet. Duckweed in the form of pellets or dried meal can been stored without difficul- ty for five or more years. Evidence suggests that it is not attacked preferentially by weevils, mice, rats, or other vermin. Duckweed as poultry and other animalfeed Feeding trials reported in the literature and carried out recently in Peru have demonstrated that duckweed can be substituted for soy and fish meals in prepared rations for several types of poultry: broilers, lay- ers, and chicks. Cultured duckweed can be used as the protein component in poultry diets. Acceptable levels of duckweed meal in the diets of layers range up to 40 percent of total feed. Duckweed- fed layers produce more eggs of the same or higher quality as con- trol birds fed the recommended formulated diets. Levels of up to 15 65 percent duckweed meal produce growth rates in broilers which are equal to those produced by control feeds. Diets for chicks, consist- ing of up to 15 percent duckweed meal, are suitable for birds under three weeks of age. Duckweed meal will almost certainly find as large a range of animal feed applications as soybean meal. Duckweed as a mineral sink Duckweed is a crop whose mi- cronutrient requirements are substantial, so much so, in fact, that waterlogged, salinized soils, which are an important constraint on irrigated agriculture worldwide, may be a favorable environment for duckweed cropping. Duckweed has the potential, thereby, to be- come the building block for integrated farming in those areas. Sev- eral types of saline environments that may be converted to duckweed cropping have considerable economic potential: (1) wa- terlogged, salinized irrigation command areas; (2) coastal wetlands; and (3) saline groundwater for irrigation or potable use. Alternative solutions to these problems are engineering-inten- sive and typically require large capital investments. Investigation to develop alternative duckweed systems to substitute for these expen- sive investments is an important area of future duckweed research. Constraints and research needs It has long been evident that duckweed has the potential to become a major protein commodity. Researchers worldwide have replicated experiments demonstrating the remarkable productivity of duckweed. Similarly, numerous studies have demonstrated the value of duckweed as a feed for poultry, fish, and other animals. However, duckweed has not yet been accepted as a commercial crop. But the major problem has been the economics of desiccation. No conventional drying technol- ogy has been able to produce a dried duckweed commodity without incurring a significant financial loss. The Mirzapur experimental program in Bangladesh represents the first effort to apply existing knowledge on duckweed growth and cultivation to develop a practical farming system. By closely tying a viable and efficient duckweed end-use (feeding fish) to duckweed production, the Mirzapur experimental program has shown that duckweed farming can be profitable. Together, these two processes represent a farming system which, in its first full production cycle, is already competitive with any crop now grown in Bangladesh. The 66 Mirzapur duckweed/carp polyculture ponds are currently the most productive non-aerated carp ponds in Bangladesh. In the Mirzapur experimental program both duckweed farming and duckweed/carp polyculture have borrowed heavily from the ex- isting literature to achieve their early success. This success has also highlighted a number of important areas for additional research. Duckweed production The most important immediate re- search priority to advance duckweed production is to determine fertilizer requirements, particularly nitrogen and trace elements. The current practice of using of urea and unrefined sea salt is clearly inadequate. Exhaustive trials are needed, first to determine nutrient requirements and then to determine efficient sources for those minerals. Farming-systems research should examine a variety of collat- eral crops which can provide efficient sun and wind buffering while maximizing total system income. Taro, for instance, works well as a buffer and provides excellent financial returns, but cannot repro- duce efficiently in water 20 to 50 cm deep. Much more work is required to understand circumstances which favor one species of duckweed over another. Although Wolf- fila species are seldom found to be dominant in the wild, it has now been successfully cultivated for two years, both singly and in com- bination, with other species. Based on current information, Wolf- fia appears to be the most productive of the three genera available in Bangladesh. Genetic improvement Little has yet been done to assess and harness genetic variance both within and among duckweed spe- cies. Studies are needed to develop strains that are more tolerant of variations in pH and temperature. Recent advances in recom- binant technology point to the possibility of developing optimized strains in the near future. By virtue of their structural simplicity and their ability to clone, the duckweed family is one of the most amenable of the higher plants to genetic engineering. Duckweed wastewater treatment Duckweed-based waste- water treatment systems have demonstrated great efficiency in treating domestic wastewater and also have done so at a net proflt. Research needs to be conducted to optimize pond design in balance 67 with the agronomic requirements for duckweed production. For ex- ample, not enough is known about the capability of duckweed to re- move heavy metals and toxins from certain types of wastewater. Answers to these questions, as well as more precise information on nutrient uptake rates, are necessary to develop standardized engi- neering guidelines for duckweed-based wastewater treatment facil- ities. Drying Duckweed will not become a traded commodity until it can be economically dried. Several solar drying methodologies have already shown considerable promise. These and other inexpensive drying technologies should be further developed to enable commer- cial-scale drying. Care should be taken to ensure that beta-caro- tene and xanthophyll are not degraded during drying. Derived products Researchers have demonstrated the ability to extract the protein fraction of wet duckweed through coagula- tion. If this process is refined and can be made cost-competitive with soy protein, the potential applications for duckweed protein are very great. High concentrations of beta carotene and xantho- phyll suggest that duckweed could become a significant source of vitamin A and pigment. Duckweed and fisheries Evidence so far suggests that duck- weed serves as a complete nutritional package for carp polyculture and can significantly increase total system productivity. The vari- ous hypotheses underlying the duckweed/carp polyculture model presented in this paper now require careful testing to explain their fundamental mechanisms. There is clearly room to optimize the model. Questions such as species mix for the polyculture, timing of harvests, length of cycle, and timing of fingerling inputs, and quan- tity of feed application require more precise answers. 68 Annexes Investment Scenarios Annex 1 estimates the profitability in Bangladesh of five-year investments in one hectare of duckweed-fed carp culture. Annex 2 analyzes costs and returns of the unit of duckweed production (0.5 hectare) necessary to support one hectare of duckweed-fed fish production. Both scenarios assume a sale price for fresh duckweed of $0.03/kg, 7 percent yearly inflation, and a 10 percent discount rate. The investment scenario for fish production assumes a sale price of $1.50/kg for fresh carp. The projected rates of return on both investments compare favorably with any alternative invest- ments in the agricultural sector in Bangladesh. Land costs for the fish culture scenario are assumed to be sig- nificantly higher ($5,000/ha versus $3,000/ha) than for the duck- weed scenario. This reflects an assumed use of marginal, unimproved land for duckweed production and use of existing, highly valued fish ponds for duckweed-fed fish production. For simplicity, both scenarios assume that all capital, includ- ing working capital, will be provided by the farmer in year zero. For that reason 'cost of capital' is not included as a line item under 're- current costs". Substitution of debt for direct investment will great- ly enhance the farmer's rate of return for each scenario. The profitability of duckweed production is especially sensitive to two factors: (1) the cost of fertilizer, and (2) the sale price of fresh duckweed. Where all fertilizer and most water are obtained from a domestic wastewater stream, the internal rate of return on duck- weed production jumps from 23 percent to 52 percent. A 30 percent increase in the price of fresh duckweed brings the internal rate of return up to 55 percent. The profitability of duckweed-fed fish production is most sen- sitive the price of fresh fish, and the cost of investment capital, but reasonably insensitive to the price of fresh duckweed. A 30 percent decrease in the price of fish reduces the internal rate of return to 16 percent, but a 30 percent increase in the price of duckweed only reduces the internal rate of return by 6 percent to 44 percent. 69 Annex 1. Investment Scenario for Duckweed-Fed Fish Production 1.0 Hectare for 5 Years COSTS (US$) Year 0 Year 1 Year 2 Year 3 Year 4 Year 5 Capital Costs Land $5,000 Pond Rehabilitabon $5,714 Water Supply $1,500 Equipment $857 Total Fixed Costs $13,071 Total Working Capital $4,200 Total Captal Requirements $17,271 Recurrent Costs Duckweed - lresh feed $3,100 $3,317 $3,549 $3,798 $4,063 Fingerlings $457 $489 $523 $560 $599 Pond Preparation $429 $459 $491 $526 $562 Water $571 $611 $654 $699 $748 Labor $712 $762 $815 $872 $933 Miscellaneous $57 $611 $654 $699 $748 Total Recurrent Costs $5,840 $6,249 $6,686 $7,154 $7,655 INCOME Sale of Fish $15,000 $16,050 $17,174 $18,376 $19,662 NET INCOME ($17,271) $9,160 $9,801 $10,487 $11,221 $12,007 CALCULATIONS (5 year investment) Internal Rate of Retum 50% Net Present Value $20,141 Break Even Point 1.8 years ASSUMPTIONS (per year) Labor 2,601 hr Water 7 0 60,000 m3 Fingerlings 20,000 i Production 10 tons 70 Annex 2. Investment Scenario for Duckweed Production 0.5 Hectares for 5 Years COSTS (US$) Year 0 Year t Year 2 Year 3 Year 4 Year 5 Capital Costs Land $1,500 Earthworks $714 Water Supply $714 Equipment $286 Duckweed Seed Stock $71 Total Fixed Costs $3,285 Total Working Capital $486 Total Capital Requirements $3,771 Recurrent Costs Fertilizer $866 $927 $991 $1,061 $1,135 Supplies $71 $76 $81 $87 $93 Bamboo, etc. $171 $183 $196 $209 $224 Water $286 $306 $327 $350 $375 Labor $548 $586 $627 $671 $718 Total Recurrent Costs $1,942 $2,078 $2,223 $2,379 $2,546 INCOME Duckweed Sale $3,142 $3.362 $3,597 $3,849 $4,119 NET INCOME ($3,771) $1,200 $1,284 $1,374 $1,470 $1,573 CALCULATIONS (5 years) Hydroponic With Wastewater Internal Rate of Return 23 % 52 % Net Present Value $2,712 $4,757 Break Even Point 2.9 years 1.8 years ASSUMPTIONS Fertilizer Urea 3,120 kg TSP 624 kg Potash 624 kg Salt 1,404 kg Water 30,000 m3/year Labor 2,000 hours Production (wet weight) 110 tons 71 Selected Bibliography Duckweed Abdulayef, D. A. 1969. "The Use of Common Duckweed as Green Feed for Chickens." Uzbekskii Biologicheskii Zournal (USSR) 13(3): 42. Buddhavarapu, L. R. and S. J. Hancock. 1991. 'Advanced Treat- ment for Lagoons Using Duckweed.' Water Environment & Tech- nology, Water Pollution Control Federation. March 1991: 41-44. Culley, D. D., E. Rejmankova, J. Kvet and J.B. Frye. 1981. "Produc- tion, Chemical Quality and Use of Duckweeds (Lemmaceae) in Aquaculture, Waste Management, and Animal Feeds." Journal of the World Mariculture Society 12(2): 27-49. Culley, D. D., and E. A. Epps. 1973. "Uses of Duckweed for Waste Treatment and Animal Feed." Journal of the Water Pollution Con- trol Federation 45(2): 337-47. Harvey, R. M., and J. L. Fox. 1973. "Nutrient Removal Using Lemna Minor." Journal of the Water Pollution Control Federation 45(9): 1928-38. Hillman, W. S., and D. D. Culley. 1978. "The Uses of Duckweed." American Scientist. 66(July): 442-51. Joy, K. W. 1969. 'Nitrogen Metabolism of Lemna minor: Growth, Nitrogen Sources and Amino Acid Inhibition." Plant Physiology 44: 842. Landolt, E. ed. 1980. "Key to Determination, Cytological Variation; Amino Acid Composition and Sugar Content." In Biosystematic Investigations in the Family of Duckweeds (Lemnaceae), vol. 1, no. 70. Publication of the Geobotanical Institute of the E.T.H. Zurich: Stiftung Rubel. _. 1986. The Family of Lemnaceae- A Monographic Study: Morphology, Karyology, Ecology. Geographic Distribution, Sys- tematic Position, Nomenclature, Descriptions. (vol. I of mono- graph). In Biosystematic Investigations in the Family of Duckweeds (Lemnaceae), vol. 2, no. 71. Publication of the Geo- botanical Institute of the E.T.H. Zurich: Stiftung Rubel. 72 Landolt, E., and R. Kandeler. 1987. The Family of Lemnaceae- A Monographic Study: Phytochemistry, Physiology, Application, and Bibliography. (vol. 2 of monograph). In Biosystematic Investiga- tions in the Family of Duckweeds (Lemnaceae), vol. 4, no. 95. Publication of the Geobotanical Institute of the E.T.H. Zfirich: Stiftung Riabel. Lueoend, A. 1983. "Growth of Duckweeds (Lernnaceae) Depending on Nutrient Supply, Especially Phosphorous and Nitrogen." t In Biosystematic Investigations in the Family of Duckweeds (Lem- naceae), vol. 3, no. 80. Publication of the Geobotanical Institute of the E.T.H. Zfirich: Stiftung Ruibel. Mbagwu, I. G., and H. A. Adeniji. 1988. "The Nutritional Content of Duckweed (Lemnapaucicostata Hegelm.) in the Kainji Lake Area, Nigeria." Aquatic Botany. 29: 357-66. Mestayer, C. R., D. D. Culley, Jr., L. C. Standifer, and K. L. Koonce. 1984. "Solar Energy Conversion Efficiency and Growth Aspects of the Duckweed, Spirodela punctata." Aquatic Botany (Amster- dam) 19: 157-70. Muzafarov, A. M. 1968. 'The Use of Common Duckweed for Feeding Domestic Birds." Uzbekskii Biologicheskii Zoumal (USSR) 12(3): 42. National Academy of Sciences. 1987. (First printing 1976). Making Aquatic Weeds Useful: Some Perspectives for Developing Coun- tries Washington, D.C. Ngo, V. 1987. "Boosting Pond Performance with Aquaculture." Op- erations Forum for Wastewater Professionals, A WPCF Publica- tion, August 1987: 20-23. Oron, G., and Dan Porath. 1987. Performance of the Duckweed Species, Lemna gibba on Municipal Wastewater for Effluent Ren- ovation and Protein Production."Biotechnology and Bioengineer- ing 29: 258-68. New York: John Wiley & Sons. Oron, G., L. R. Wildschut, and Dan Porath. 1984. "Waste Water Re- cycling by Duckweed for Protein Production and Effluent Reno- vation." Water Science Technology (Amsterdam) 17: 803-17. Oron, G., and Hans Willers. 1989. "Effect of Wastes Quality on Treatment Efficiency with Duckweed." Water Science Technology (Amsterdam) 21: 639-45. 73 Porath, D., G. Oron, and G. Granoth. 1985. "Duckweed as an Aquat- ic Crop: Edible Protein Recovery, Production and Utilization." In Proceedings of the Fifth Symposium on Agricultural Wastes (Publi- cation 13). St. Joseph, Michigan: American Society of Agricultural Engineering. Rusoff, L. L., W. W. Blakeney, and D. D. Culley. 1980. "Duckweeds (Lemnaceae Family): A Potential Source of Protein and Amino Ac- ids." Journal of Agricultural and Food Chemistry 28: 848-50. Rusoff, L. L., D. T. Gantt, D. M. Williams, and J. H. Gholson. 1977. "Duckweed - Potential Feedstuff for Cattle." Journal of Dairy Sci- ence6(sl): 161. Rusoff, L. L., S. P. Zeringue, A. S. Achacoso, and D. D. Culley. 1978. "Feed Value of Duckweed (An Aquatic Plant: Family Lem- naceae) for Ruminants." Journal of Dairy Science 61 (s 1): 18 1. Saline Agriculture: Salt Tolerant Plants for Developing Countries. 1990. Washington, D.C.: National Academy Press. Haustein A. T., W. R. Gilman, and P. W. Skillicorn. 1990. "Duck- weed, A Useful Strategy for Feeding Chickens in Third World Countries: Performance of Layers Fed With Sewage-Grown Lem- naceae." Poultry Science 69: 1835-44. Stanley, R. A., and C. E. Madewell. 1976. "Thermal Tolerance of Lemna minor." Muscle Shoals, Alabama: Tennessee Valley Au- thority. Truax, R., D. D. Culley, M. Griffith, W. Johnson, and J. Wood. 1972. "Duckweed For Chick Feed?" Louisiana Agriculture 16(1): 8-9. Wolverton, B. C., R. M. Barlow, and R. C. McDonald. 1976. "Appli- cation of Vascular Aquatic Plants for Pollution Removal, Energy and Food Production in a Biological System." In J. Torbier, and R. W. Pierson, eds., Biological Control of Water Pollution 141-49. University of Pennsylvania Press. 74 Fish culture Bardach, J. E., J. H. Ryther, and W. 0. McLarney. 1972. "Aquacul- ture: The Farming and Husbandry of Freshwater and Marine Or- ganisms." New York: John Wiley & Sons. Blackburn, R., and D. L. Sutton. 1971. "Growth of the White Amur (Ctenopharyngodon idella Val.) on Selected Species of Aquatic Plants." In Proceedings of the European Weed Research Councils 3rd International Symposium on Aquatic Weeds. Buck, H., R. J. Baur, and C. R. Rose. 1978. 'Polyculture of Chinese Carps in Ponds with Swine Wastes." In Smitherman, et al., eds., SyTnposium on Culture of Exotic Fishes 144-55. Auburn, Ala- bama: Ameiican Fisheries Society. Huismann, E. A. 1979. "The Culture of Grass Carp (Ctenopharyn- godon idella Val.) Under Artificial Conditions." In Flnfish and Fishfeed Technology. Halver, J. E., and K. J. Tiews, eds., 491- 500. Berlin: Heinemann Verlagsgesellschaft. Jauncey. K.. and A. Matty. 1979. "Mirror Carp-Fast Grower With Room for Expansion." Fish Farmer 2(5): 29. Maddox, J. J., L. L. Behrends, C. E. Maxwell, and R. S. Pile. 1978. "Algae-Swine Manure System for Production of Silver Carp, Big- head Carp and Tilapia." In Smitherman, et al., eds., Symposium on Culture of Exotic Fishes. Auburn, Alabama: American Fisher- ies Society. Murshed, S. M., S. N. Roy, D. Chakraborty, M. Randhir, and V. G. Jhingran. 1977. 'Potential and Problems of Composite Fish Cul- ture Technology in West Bengal." Bulletin of Central Inland Fish- ery Research Institute (Barrackpore, India) 25: 11. Pullin, R. S. V., T. Bhukaswan, K. Tonguthai, J. L. McLean, eds. 1988. Second Symposium on Tilapia in Aquaculture. Manila: In- ternational Center for Living Aquatic Resources Management. Schroeder, G. L. 1979. "Microorganisms As the Primary Diet in Fish Farming." 14/15: 373-83. Bundesforschungsamt fur FYscheret Hamburg, Germany. Shireman, J. V., D. E. Colle, and R. G. Rottman. 1977. "Intensive Culture of Grass Carp, Ctenopharyndogon idella, in Circular Tanks." Joumal of Fish Biology 11(3): 267-72. 75 Shireman, J. V., 1978. 'Growth of Grass Carp Fed Natural and Pre- pared Diets Under Intensive Culture." t Journal of Fish Biology 12(5): 457-63. Shreenivasan, A. 1979. "Fish Production in Some Hypertrophic Ec- osystems in South India." Devopment Hydrobiology (Nether- lands) 2: 43-50. Stanley, J. G. 1974. "Nitrogen and Phosphorus Balance of Grass Carp, Ctenopharyngodon idella, Fed Elodea, Egeria densa." 7Tan- scripts of the American FYshery Society 103(3): 587-92. 76 This publication is based primarily on a study per- formed at the Mirzapur Experimental Duckweed Site by The PRISM Group of Columbia, Maryland, U.S.A. The study de- scribes current knowledge about farming aquatic plants of the family Lemnaceae, the common duckweeds, their poten- tial as a protein-rich animal feedstuff, and their value as a low cost, low energy wastewater treatment technology. The World Bank . _~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4 Headquarters 1818 H Street. N.W. Washington, D.C. 20433, U.S.A. Telephone: (202) 477-1234 Facsimile: (202) 477-6391 Telex: WUI 64145 WORLDBANK RCA 248423 WORLDBK Cable Address. INTBAFRAD WASHINGTONDC European Office 66, avenue d'lena 75116 Paris, France Telephone: (1) 40.69.30.00 Facsimile: (1) 40.69.30.66 Telex: 640651 Tokyo Office Kokusai Building 1-1 Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan Telephone: (3) 3214-5001 Facsimile: (3) 3214-3657 Telex: 26838 12067 AGR 100 F 0-8113-2067-X :DUCKWEED AQUACULTURE 400000005102 6.95 t; ~ IB 0-8213.......2067-X....