You are on page 1of 8

Bioresource Technology 73 (2000) 13±20

In vitro selection of duckweed geographical isolates for potential use in swine lagoon e‚uent renovation
B.A. Bergmanna,*, J. Chengb, J. Classenb, A.-M. Stompa
b a Department of Forestry, Campus Box 8002, Raleigh, USA Department of Biological and Agricultural Engineering, Campus Box 7625, North Carolina State University, Raleigh, NC 27695, USA

Received 15 December 1998; received in revised form 24 September 1999; accepted 30 September 1999

Abstract Plant-based systems for nutrient sequestration into valuable biomass have the potential to help avoid the environmental problems associated with the disposal of large volumes of animal waste. The objective of this study was to select superior duckweed (Lemnaceae) genotypes for the utilization of nutrients in animal wastes. A two-step protocol was used to select promising duckweed geographic isolates to be grown on swine lagoon e‚uent. Forty-one geographic isolates from the worldwide germplasm collection were used in an in vitro screening test, because they were noted to be fast-growing genotypes during routine collection maintenance. In vitro screening was accomplished by growing geographic isolates on a synthetic medium that approximated swine lagoon e‚uent in terms of nutrient pro®le, total ionic strength, pH, and bu€ering capacity. Large di€erences among geographic isolates were observed for wet weight gain during the 11-day growing period, percent dry weight, and percent protein in dry biomass. Total protein production per culture jar di€ered 28-fold between the most disparate of the 41 geographic isolates and was the variable used for selection of superior geographic isolates. The challenge of eight of the 41 geographic isolates with full-strength swine lagoon e‚uent in the greenhouse led to the selection of three that are promising as genotypes to be grown on lagoon e‚uent. Ó 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Duckweed; Growth; Lemnaceae; Lemna gibba; Lemna minor; Protein production; Spirodela punctata; Swine waste; Wola; Wolella

1. Introduction Enormous quantities of nutrients are present in animal wastes produced at modern, large-scale animal production facilities, a situation that has created global environmental concern (Williams, 1995). Technologies are needed for e€ective utilization of waste stream nutrients so that they do not contribute to surface and ground water contamination. Alternative plant-based systems for sequestering nutrients into biomass potentially may avoid the environmental problems associated with disposal of animal wastes. Land application of animal waste is common and relatively successful in the midwestern US where deep, permeable soils and integration of feed crop and animal production allow nutrient recovery from the manure stream via land application systems with feed crops such as corn. However, in the southeastern US, swine farms are not associated with large tracts of land, corn and

Corresponding author.

soybean yields are lower than in the Midwest, and the utilized crops are selected for nutrient uptake rather than usefulness (such as feed grains). These factors result in the transformation of animal waste nutrients from one problematic form to another, i.e. large amounts of low value biomass that is stockpiled in nutrient rich areas. Build-up of some waste constituents in the soils to which animal waste has been applied over an extended period of time has been documented. This is particularly true in the case of poultry litter, which can lead to soil accumulation of excess phosphorus (Heathman et al., 1995; Chapman, 1996) and copper and zinc (Mitchell et al., 1992; van der Watt et al., 1994). Elevated nitrogen loading rates can increase the NO3 -nitrogen levels in forage biomass to a level that raises toxicity concerns (Harvey et al., 1996) because excess nitrate is reduced to nitrite in the digestive tract of grazing ruminants, and the resulting excess nitrite limits blood oxygen transport (Ball et al., 1996). Treatment technologies that are e€ective only during a given cropÕs active growing season require storage of

0960-8524/00/$ - see front matter Ó 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 0 - 8 5 2 4 ( 9 9 ) 0 0 1 3 7 - 6

a logical starting point for research with duckweed is improvement with respect to desired traits.. 1973. Haustein et al.15 g/l citric acid was used as a bu€ering agent. maize.. 1992. Mbagwu and Adeniji. total ionic strength. In vitro screening An in vitro protocol was used for the selection of promising geographic isolates to be grown on swine lagoon e‚uent in the greenhouse. This step has been accomplished by the transfer of the worldwide duckweed germplasm collection from Dr. and ®sh (Shireman. 1992) and 54. 2. Once such a match is made. Rodriguez and Preston. 1975. 20. This was accomplished by completely covering the surface of a water-®lled 10 ml beaker with fronds before placing them into the larger culture vessels and was done to reduce bias by di€ering frond sizes among geographic . the genetic make-up of the population used for nutrient sequestration should stay relatively constant as the vast majority of progeny are produced through asexual reproduction from mother fronds. 1979. 1993) is a valuable trait. The 41 geographic isolates represented 12 species.. the tendency of duckweed to assimilate NH‡ 4 eciently (Monselise et al... This creates the opportunity to select superior geographic isolates and to match suitable geographic isolates to speci®c environments and/or waste streams. at least one species from each of the four genera within Lemnaceae (Table 1).0. This was done because in vitro conditions permitted the testing of a large number of genotypes in a relatively small area. 1. and the pH after autoclaving was 7. The collection consists of nearly 1000 geographic isolates belonging to 36 species within the four genera in Lemnaceae: Lemna. / Bioresource Technology 73 (2000) 13±20 large volumes of waste at year-round animal production facilities at a signi®cant cost to the producer. Wolella.. Oron.0 cm tall) with polypropylene caps (Magenta Corp. The next step is to select genotypes that exhibit desirable traits.0 cm outer diameter ´ 7. swine (Leng et al. The carbon source in SAM was 3. 1980.. E€ective removal of nutrients from wastewater through duckweed production has been demonstrated (Harvey and Fox... Forty-one geographic isolates were used. 1988. 1988. Leng. The objective of the research presented was to identify one or more duckweed geographic isolates especially suited to growth on swine lagoon e‚uent. Ruso€ et al. Zurich. Glass baby food jars (6. Buddhavarapu and Hancock. Oron et al. the growth chamber provided the uniformity needed to evaluate genotypes without the confounding in¯uence of ¯uctuating environmental conditions. It is logical to focus on duckweed species (Lemnaceae) for use in aquatic systems designed for recovery of nutrients in animal wastes. In general. 1988) and has been shown to be an adequate substitute for high protein supplements such as soymeal in feed for poultry (Truax et al.14 B... protein production. 1994). 1990). These would compliment the geographic isolates shown to be most ecient for biomass/protein production during the heat of the summer. Leng et al... When applied to nutrient utilization from animal waste. pH. Switzerland) to the Environmental Biotechnology Laboratory at North Carolina State University.. 1997). 1990. Other features that make duckweed an attractive target for nutrient utilization systems include its great genetic diversity and the fact that it clones itself. Jauncey and Matty. 1991). 1978. 1995. Also. Sutton and Ornes. duckweeds grow over a longer period compared to crops with a determinate growth habit (e. The length of the e€ective wastewater renovation period for a duckweed-based system might be extended by selecting geographic isolates that perform well in the early spring and late autumn. Medium was autoclaved at 121°C for 18 min. Eight of the geographic isolates were native to the United States The culture medium was designated swine arti®cial medium (SAM) as it was formulated to closely resemble the nutrient pro®le. 1990. 1973. 1990). 1994). the ability to proliferate at high or low temperatures. Gaigher et al. dairy cattle (Culley and Epps. 1996. 1995. An obvious end-use for dried duckweed is as an animal feed supplement. Dry matter production is high compared with other crops. Bergmann et al. Methods 2. 1984. e. The wide geographic distribution of populations within the 36 species within Lemnaceae suggests that signi®cant variation will be detected among genotypes for traits such as the ability to grow well on di€erent types and concentrations of wastewater. Elias Landolt (ETH. Ruso€ et al.A. and suitability of the biomass produced for di€erent end uses. as it typically contains between 24% and 45% protein (Edwards et al. Oron. Hang et al.0% sucrose. Van et al. Wola. Spirodela. soybeans). chosen from the collection based on the observation that they exhibited rapid growth during routine culture maintenance. When considered as a new crop. Each jar was inoculated with fronds from one of the 41 geographic isolates such that 35% of the medium surface area was covered. 1997. 1978. particularly important in cases such as duckweed for which asexual propagation will be used.g.8 t/ha/yr for Lemna gibba (Oron et al.1. 1987) and in preference to NOÀ 3 (Monselise and Kost. and buffering capacity of swine lagoon e‚uent commonly found in North Carolina (Table 2). 1977. Generally.. Hassan and Edwards. 1972. Chicago IL) served as culture vessels.4 t/ha/yr reported for Spirodela polyrrhiza (Edwards et al. the ®rst step in a crop improvement e€ort is to acquire a diverse genetic base upon which to build the program. 1991.g. Ray and Das. 1992.. Ruso€ et al.

141 1. Analysis of variance were .0 Bu€ered SAMb 26. The variable of most importance when considering the practical implementation of a system is the quantity of biomass/protein produced when a given surface area is covered at the outset of a given culture period.8.B. 7441 (Norway). 8626 (Denmark).0% sucrose and adjusted to pH 5.40 1.2.63 0. Frond fresh weight was recorded for each jar at the start and end of the culture period.82 0. The culture period was 11 days.063 0. 7784 (Ethiopia). 8230 (Malaysia). 8627 (Denmark). 8678 (India) 7501 (Ohio USA). 8731 (New Zealand). 8197 (China). Fronds were blotted dry with paper towels and placed into an incubator maintained at 37°C for 72 h prior to recording frond dry weight for each jar. 9008 (Columbia) 7720 (Virginia USA) 8821 (Argentina). (Colorado USA). the duckweed geographic isolates had been precultured for 2 weeks on Schenk and Hildebrandt (1972) medium (SH) supplemented with 3.016 0. These two values were used to calculate the fresh weight gain during the culture period. Thus. 7741 (Italy).35 0. 9176 (Bolivia) Table 2 Ionic concentrations and pH of swine lagoon e‚uent and the bu€ered swine arti®cial medium formulated to approximate it for use in in vitro trials Ion (mM) NH4 NO3 P K Ca Mg Cl Fe S Na B Mn Zn Cu Mo Co pH a Ionic concentration (mM) Swine lagoon e‚uenta 26. 8405 (France). Statistical analysis Statistical analyses were accomplished using SAS (SAS Institute.5 0.82 0.16 12.4 3. 8829 (Argentina) 7178 7401 8229 7488 (Argentina) (Poland). Prior to culture initiation.65 8.5 0.4 12. 8654 (China). Each geographic isolate was represented by three jars. isolates.0 An average of 715 observations were used to calculate each of the ion concentrations in swine lagoon e‚uent from North Carolina farms. Frond protein content was determined using the Lowry protein determination procedure (Stoscheck. 1990). there was a total of six jars per geographic isolate and a total of 246 jars.16 12. 8715 (Malaysia). 7968 (Alabama USA). b SAM ˆ swine arti®cial medium.063 0. 2.019 0. 7558 (California USA).15 g/l citric acid before autoclaving. Cary.015 3. 6861 (Italy). 7776 (Australia) Number of geographic isolates 7 6 7 3 1 2 1 7 2 1 1 3 15 7267 (Australia) 8776 (California USA) 7201 (Argentina). / Bioresource Technology 73 (2000) 13±20 Table 1 Duckweed geographic isolates used for in vitro screening on swine arti®cial medium Genus Lemna Species aequinoctialis gibba minor minuta obscura valdiviana Spirodela intermedia polyrrhiza punctata Wola Wolella australiana lingulata oblonga Geographic isolates 7255 (Ghana). 8683 (Kenya) (Florida USA). Replicates were placed on separate shelves in the growth chamber. 8745 (Canada) 6600 (California USA). Previous experiments indicated biomass production would likely be biased in favor of geographic isolates with large fronds when inoculating with a ®xed number of fronds and in favor of geographic isolates with small fronds when inoculating with a ®xed initial fresh weight. 8816 (Argentina).015 3.A. 9049 (Zimbabwe).65 8.00042 7.05 1. and the experiment was repeated once.019 0.141 8. 9076 (Zimbabwe) G3 (unknown). Bu€ering was accomplished by addition of 1. Total protein production for each jar was calculated by multiplying frond fresh weight gain by frond percent dry weight and multiplying the product by percent protein content on a dry weight basis. NC). (Malaysia).047 0. 6747 (California USA).06 7.016 0.00021 0. The experiment was conducted as a completely randomized design.59 7.00042 7. 8744 (Albania).047 0. Jars were placed into a 23°C growth chamber with a 16-hour photoperiod and a photosynthetic photon ¯ux density of 40 lmol/m2 /s provided by wide spectrum ¯uorescent tubes. 8240 8652 (China). Bergmann et al.00021 0.

protein content on a dry weight basis. The independent variables were placed in the following order in the statistical model: replicate. Day-time air temperatures ¯uctuated between 19°C and 34°C.A.01) among geographic isolates when an equal surface area was used to inoculate each vessel.01) was found to be as great or greater than that of species for all variables. Bergmann et al. In no other case was the in¯uence of replicate. Water lost to evaporation was replaced with tap water twice a week. and protein content. unpublished data): Lemna minor geographic isolates 8627 and 8744 which were ranked 14 and 30. and 35th for protein content. species. jar within replicate. Protein content based on dry weight di€ered (p < 0. but water temperature was maintained at 25 Æ 2°C. Once it was determined from the general linear models procedure that each of the observed variables varied signi®cantly among species. percent protein on a dry weight basis. Geographic isolate rankings di€ered markedly among fresh weight gain. sixth for percent dry weight.01) among geographic isolates with a 9-fold di€erence exhibited between the geographic isolates with the highest and lowest concentrations (Table 4). One tub per geographic isolate was used in each of two sequential replicates. DuncanÕs critical range tests were used to separate species means for dependent variables. Plastic tubs were placed into insulated boxes with bottom heat in the greenhouse. Lemna aequinoctialis 8230 was ranked ®rst for fresh weight gain. and growth chamber conditions were the same as described previously. DuncanÕs critical range tests were used to separate geographic isolate means for dependent variables after the general linear models procedure showed that each of the observed variables varied signi®cantly among geographic isolates. Data from all 41 geographic isolates were used in analyses to focus on geographic isolate as a variable because of its clear in¯uence on all variables.05) for any variable. Pearson correlation coecients were calculated between pairs of independent variables using the 41 geographic isolate means for each variable: fresh weight gain. The eight geographic isolates were grown in the growth chamber in large ¯asks to produce enough fronds to inoculate large tubs in the greenhouse. / Bioresource Technology 73 (2000) 13±20 done using a data set with only the six species (33 geographic isolates) which were each represented by more than two geographic isolates in order to examine the in¯uence of species on the observed variables.01) among geographic isolates with a 3-fold di€erence between the highest and lowest geographic isolates (Table 4). The variation among geographic isolates within species (p < 0. Qualitative observations of frond health were the only recorded data during the 4-week growing period. Two additional geographic isolates were included based on the fact that ecient genetic engineering protocols have been developed for them (Stomp and Rajbhandari.01) among geographic isolates with an 11-fold di€erence between the geographic isolates exhibiting the most and least growth (Table 4). Growth of selected geographic isolates on swine lagoon e‚uent in the greenhouse The six duckweed geographic isolates (i. For example. Results 3. and replicate by geographic isolate interaction. The 24 l non-diluted swine lagoon e‚uent (taken from the NCSU Field Laboratory) placed into each tub created a 36 ´ 80 cm2 surface area. and total protein production (Table 3). Fresh weight gain during the culture period di€ered (p < 0. The independent variables in the analyses of variance were placed in the following order: replicate. Percent dry weight for fronds di€ered (p < 0. and total protein production. fresh weight gain.8. The medium was SH supplemented with 3.01): initial fresh weight. 15% of the original 41) that were most highly ranked on the basis of absolute protein production from the in vitro screening experiment were used to observe growth and performance on non-diluted swine lagoon e‚uent in the greenhouse. jar within replicate.3. percent dry weight.e. As expected. percent dry weight. respectively.03) in the case of percent dry weight was considered an artifact. 3. and geographic isolate within species followed by the one-way interactions of replicate with species and with geographic isolate within species. A sucient number of fronds was placed in each tub to completely cover the water surface at the start of each replicate. or replicate by geographic isolate interaction signi®cant.0% sucrose and adjusted to pH 5. percent dry weight. After the strong in¯uence of geographic isolate on the dependent variables was revealed. In¯uence of replicate and one-way interactions with replicate were not signi®cant (p > 0. The test was conducted during late spring and early summer without supplemental lighting or shading. geographic isolate. a second set of analyses was done using the entire data set to focus on geographic isolates.16 B. Lemna gibba G3 was ranked 27th and 37th for fresh weight gain and . 2. In vitro screening Forty-one duckweed geographic isolates were grown in vitro on bu€ered SAM for 11 days. the initial frond fresh weights varied (p < 0. Data from the six species that were each represented by more than two geographic isolates showed that species in¯uenced all the observed variables (p < 0.1. Non-diluted swine lagoon e‚uent was used to provide the greatest challenge to the eight geographic isolates. The signi®cant in¯uence of jar within replicate (p ˆ 0.

7 13.04 mg/jar for Wolella oblonga 8817 to 28.22 0.9 d a b c d c Total protein production (g)b 0.20 0.A.6 15. Geographic isolates were highly variable (p < 0.21 0.17 0.4 a e c e b d Protein content (%) 6. Growth of selected geographic isolates on swine lagoon e‚uent in the greenhouse Characteristics of the swine lagoon e‚uent used for the greenhouse experiment are shown in Table 6.6 18.0 Dry weight (%) 12.18 0.B.5 37. Bergmann et al. These six geographic isolates tend to be relatively highly ranked based on fresh weight gain.5 10. but was ®rst in protein content.8 9.49.2 17.20 0.2 18.17 0.6 Dry weight Rank 9 29 21 5 38 16 % 22. The six most highly ranked geographic isolates according to total protein production are shown in Table 5.01) was found between fresh weight gain and total protein production (Fig. p < 0. were signi®cant.9 See text for calculation.23 0.1 7. the variable ``total protein production'' was calculated to yield a single variable by which the most productive geographic isolates were selected.05 level for 41 means. and protein content on a dry weight basis for the most disparate of 41 duckweed geographic isolates screened over an 11-day period growing in vitro on bu€ered swine arti®cial medium Geographic Isolateb Lowest ranked Critical rangec Highest ranked Mean of all 41 a b Fresh weight gain (g/jar)a 1.0 15. 3.6 6. / Bioresource Technology 73 (2000) 13±20 17 Table 3 Performance of duckweed species (represented by three or more geographic isolates each) grown in vitro on bu€ered swine arti®cial medium for 11 daysa Species Lemna aequinoctialis Lemna gibba Lemna minor Lemna minuta Spirodela polyrrhiza Wolella oblonga a b Fresh weight gain (g)b 8.02 c a b c c d Values within a column followed by the same letter are not di€erent at the 0.47.7 8. None of the geographic isolates grew rapidly. Table 4 Fresh weight gain.4 19. Total protein production was not in¯uenced by replicate.4 9.4 6.3 13. c According to DuncanÕs critical range test at the 0. 1).01).5 8.9 8. percent dry weight.24 0. According to DuncanÕs critical range test at the 0. percent dry weight. ranging from 1.5 18.4 19.5 See text for calculation.5 Protein content (%) 2.87 mg/jar for Spirodela punctata 7776.5 15.2 10.1 6.01).05 level. p < 0.21 0. jar within replicate.16 0.8 20. p < 0. but their Table 5 The six duckweed geographic isolates of 41 tested ranked most highly based on total protein productiona over an 11-day period growing in vitro on bu€ered swine arti®cial medium and rankings and values for their fresh weight gain.2.64. and strains with high protein content tended to have high total protein production (r ˆ )0. .8 ab c c a b d Percent dry weight 27.5 Protein content Rank 17 2 14 24 6 30 % 11. respectively.7 15.8 9. but there was no consistent trend with regard to ranking according to percent dry weight or protein content (Table 5). Lowest and highest ranked clones are di€erent for each variable.1 1.04 Fresh weight gaina Rank 7 21 14 9 5 2 g 9.7 20. For this reason.8 12.3 23.2 1.6 6. and protein content on a dry weight basis Species ± geographic isolate Spirodela punctata ± 7776 Lemna gibba ± 8678 Lemna minor ± 7501 Spirodela punctata ± 7488 Lemna obscura ± 7720 Lemna aequinoctialis ± 8715 Critical rangeb a b Total protein production Rank 1 2 3 4 5 6 g 0. The closest association (r ˆ )0. Values correspond to the appropriate number of intervening means for a given variable.2 7.2 16.05 level according to DuncanÕs critical range test.1 1. or replicate by geographic isolate interaction. calculated using geographic isolate means.01) in total protein production.9 13. percent dry weight.6 1.0 2.3 7.4 1.4 2. Three of the six correlations between pairs of dependent variables. High percent dry weight was associated with relatively low protein content (r ˆ )0.3 16. See text for calculation.

22 mM 3. in the in vitro screening experiment. n selected for in vitro performance followed by good greenhouse growth. 4.54 mM 2. / Bioresource Technology 73 (2000) 13±20 experiment.004 mM 0.92 mM 4. and total protein production over a de®ned growing period.01 mM 0. 1999).60 mM 1. The spread for the most disparate Lemna minor geographic isolates according to total protein production was from 3rd to 30th. These three geographic isolates originate from widely divergent points. 3 ± L. However. India. Similar performance di€erences were reported when 11 duckweed geographic isolates were grown on partially treated domestic sewage (House et al. and Spirodela punctata 7776 remained relatively healthy without noteworthy multiplication. fronds of Lemna minor 8627 exhibited about 50% survival (without noteworthy multiplication) despite its ranking of fourteenth in the in vitro screening . h selected for in vitro performance followed by poor greenhouse growth. Discussion The promise for selecting outstanding duckweed genotypes for use in waste nutrient utilization systems was demonstrated.e.64.71 mM 9. Qualitative observations were uniform between replicates and led to the following ranking of the eight geographic isolates according to health during 4 weeks on non-diluted swine lagoon e‚uent in the greenhouse: 1 ± L. The large in¯uence of geographic isolate on the observed traits suggested that it would be practical to focus attention on the geographic isolate rather than the species level.41% K Cl Na Ca Mg Cu Zn Fe pH 6. 5 ± L. For example. punctata 7488. d not selected in vitro and exhibited good greenhouse growth. s not selected in vitro and exhibited poor greenhouse growth. The single trait total protein production resulted from the combination of three other traits and was used for selection in this work.06 mM 0. This approach is supported by observations of geographic isolates that respond relatively well or poorly for a given variable compared to the species to which it belongs. The remaining geographic isolates of the six selected in the ®rst experiment did not tolerate non-diluted swine lagoon e‚uent. while Lemna gibba produced the most total protein as a species during the in vitro trial. responding with slow senescence of the majority of fronds.84 responses di€ered markedly as judged by non-quanti®ed. The rapid senescence of Lemna minor 8744 fronds was not surprising given its ranking of thirtieth in the ®rst experiment. Geographic isolate symbols: + tested in vitro only.. care must be taken in choosing selection criteria since selection can be Fig. Fresh weight gain and total protein production by 41 duckweed geographic isolates during an 11-day growing period in vitro on bu€ered swine arti®cial medium. 2 ± S.20% 42. 6 ± L.83 mM 0. percentage dry weight. 1.0001. punctata 7776.01 mM 2.006 mM 0. its geographic isolates ranged in rank from 2nd to 18th among the 41 tested. minor 8744.A. respectively. visual estimates. Australia. gibba 8678. The fact that the three promising geographic isolates in the present work are in two genera and that each is within a di€erent species points to the usefulness of selecting directly at the geographic isolate level rather than starting at the genus or even species level. However.67 mg/l 0. These geographic isolates were ranked second and ®rst. Clear di€erences were found among species and geographic isolates for traits that are important for production of valuable biomass on swine lagoon e‚uent: fresh weight gain.18 B. minor 8627. Bergmann et al. it would be wise to screen a large number of duckweed isolates endemic to a relatively limited geographic area given the concern for the potential invasiveness of exotic isolates.54 mM 0. and Denmark. Table 6 Characteristics of unmodi®ed swine lagoon e‚uent used to grow eight duckweed geographic isolates in the greenhouse TKN NH4 NO3 NO2 TP OPO4 TOC COD TS VS 18. obscura 7720 and L. protein content on a dry weight basis. Lemna gibba 8678 appeared to tolerate the lagoon e‚uent better than the other geographic isolates and maintained healthy fronds that multiplied slowly. which indicates that selection from a genetically diverse duckweed collection for a given trait could a€ord signi®cant improvement compared to the duckweeds that happen to grow locally. i. p ˆ 0. Response di€erences among geographic isolates were expected given that 41 were used from 12 species within the four Lemnaceae genera.33 mg/l 1286. minor 7501 and S.042 mM 7.46 mM 421. If a selected duckweed is intended for use in a non-con®ned system. Pearson correlation coecient for the two variables based on geographic isolate means (n ˆ 6): r ˆ 0. aequinoctialis 8715. 4 ± L. The two geographic isolates included because they can be eciently genetically engineered di€ered in response to non-diluted swine lagoon e‚uent.

15 N nuclear magnetic resonance study of ammonium-ion assimilation by Lemna gibba L. a useful strategy for feeding chickens: Performance of layers fed with sewage grown Lemnaceae species..J.A. Agric. A. P. D. Res. 1835±1844. 3.. selection of traits in sequence could be used to de®ne a select group of geographic isolates.. 1991.C. A.M.. 40.A. Chao.I.. House. Di€erent ammonium-ion uptake. Sci. Birmingham. Kost. Stamboli. C. Economic considerations in wastewater treatment with duckweed for e‚uent and nitrogen renovation.. Hassan. J. Hannan. Evaluation of duckweed (Lemna Gibba) as feed for tilapia (Oreochromis niloticus cross Oreochromis aureus) in a recirculating unit.N.B....M. The goal of the continued research will be to de®ne a continuous-¯ow swine lagoon e‚uent treatment system that simultaneously produces biomass of value. 235± 244. V. Nelson.S. and temperature ¯uctuations are some of the variables that would signi®cantly in¯uence duckweed biomass and protein production and that di€er greatly between the growth chamber and in situ systems. 357±366. Poultry Sci. V. J. M. Epps. P.G.. 10. Compensatory growth in broiler chicks fed on Lemna gibba. Finally. S. Auburn Univ Press.. the top six from among those 11 geographic isolates were selected according to percent dry weight.. Windham. and water quality from bermudagrass pastures fertilized with two levels of nitrogen from swine lagoon e‚uent. R. Factors a€ecting the utilization of `poor quality' forages by ruminants particularly under tropical conditions. 1995. 74. 1990..B. M. 1995. C. Fert... M. and L. V. Nitrogen digestion and metabolism in Mong Cai pigs fed sugar cane juice and di€erent foliages as sources of protein. H. 1992. Heathman. 36.B. Greenhouse and ®eld tests utilizing a large number of geographic isolates would be cumbersome and costly compared to the in vitro screening method used here. Edwards. Among the six selected via this method were the top four selected by using the single trait total protein production. References Ball.. Oron...D. ranked 14th in that method but shown to be promising for growth in the greenhouse on swine lagoon e‚uent. 29. 2nd ed.. Brit. Water Pollut.. Mitchell. Leng. minor 8627 because they grew on non-diluted swine lagoon e‚uent in the greenhouse after having been selected for total protein production in vitro on medium that closely resembled typical swine lagoon e‚uent. 1992. 1984. Nutr. 457±464.. Ly. Res. J. 1993. 6±8 October. Soil and solid poultry waste nutrient management and water quality. J.. D.. Res.. R. 1928±1938. 1996. G. 1990.A.) in the Kainji Lake area. Bell. C. two of which were among the six selected using the single combined trait. N.H.A. 1973. 122. A.A. Advanced treatment for lagoons using duckweed. Livestock Research for Rural Development 7 (1).. Hang. J. M. B.. D. G. Light level. A. 341±345. Matty. I.A. Frederick. R. 1997.H.. A. Aquaculture 41.. Bioresource Technol. 329±335.. Control Fed. R.S.L. Porath. Mbagwu. / Bioresource Technology 73 (2000) 13±20 19 accomplished via several methods. Cultivation of duckweeds on septage-loaded earthen ponds. Bergmann et al... S. Buddhavarapu. Eng. One of the two that were changed from the total protein production selection method was Lemna minor 8627. 45 (2).W. D.. E€ects of long-term litter application on coastal plain soils. Guevara. Lace®eld.. Ecol. E.P. 285±289..B.. 315±326.. K. Monselise. P. The ranking of these 41 geographic isolates would likely change if all were tested in the greenhouse or ®eld compared to rankings in vitro..J.. 1990. Tol. one could focus on the set of geographic isolates that emerges when only those are selected that are in the top 50% of all desired traits. USA. 68. P. Gilman. Haustein. Planta 189. P. Control Fed. R. Skillicorn. 1987.. 12 (1&2). March. Poore.. Caution must be used in assuming that the types and magnitudes of variation in a controlled in vitro system would be seen in a ®eld system. including the two that performed well on non-diluted swine lagoon e‚uent in the greenhouse.. I. In: Blake et al.J. Gilman. 109±117. Edwards. 45 (9). steer performance. 277±303.A. GA. Adeniji. Hoveland. Kost. Chapman. punctata 7776. A. Combining constructed wetlands and aquatic and soil ®lters for reclamation and reuse of water. J. AL.. Bergmann. J. 1994. J. Zublena. Hancock. J. D. Porath. M. Bot. J. Vergara. Potash and Phosphate Institute. pp. Haustein. E. Water Pollut. another factor that will in¯uence duckweed performance.R.L. Aquaculture 104. Barker. greenhouse and ®eld trials will continue with L. Performance of broiler chickens fed diets containing duckweed (Lemna gibba). L. Leng. Poultry Sci. Gaigher. Duckweed.. G.. Smith. Control Fed. C... Fish Farmer 2 (5)..B. Southern Forages. 75.. New Phytol.G. 862±866.S.P. Pacharaprakiti... The nutritional content of duckweed (Lemna paucicostata Hegelm.. Guevara.. 165±173. Granoth. 1996. J. D.. Uses of duckweed for waste treatment and animal feed.T. L. 1999. Skillicorn. 1973. A. D. Nutrient removal using Lemna minor. 167±193. Forage characteristics. Batches of swine lagoon e‚uent will vary in nutrient levels. Water Pollut. Land application of poultry litter and water quality in Oklahoma. Robinson. 337±347. A. 1992. R. Gilman.. . Control Fed. R..W.S.. Gilman. Haustein. 29. S. USA. This was done with the 41 geographic isolates by selecting the top 21 with regard to fresh weight gain (the trait most closely associated with total protein production) and then selecting the top 11 of those 21 with regard to percent protein (important in choosing geographic isolates of potentially greater value). gibba 8678. Diaz. Culley. Proceedings of National Poultry Waste Management Symposium Committee. Fox. 40. 1979. G... For example. Stomp. R.. Water Pollut. 383±390. Rev. 62 (5). D.. Rodriguez. 1988. S..T.. J.H. Gastanaduy. Lai. Nutr. J. Harvey. temperature. (Eds. Aquat. Therefore. F. Jauncey. Harvey. Water Environment & Technology. V. Nigeria.. E. 27±38. Gastanaduy...M. 69. Baltikauski. Monselise. 692±696.).. Animal Sci.. Skillicorn. Hassan. Livestock Research for Rural Development 9 (2). C.H. 1991. Duckweed ± a potential high protein feed resource for domestic animals and ®sh.. Mirror carp ± fast grower with room for expansion. 41±44. Sharpley. H... Vergara. S.N.. 1996..C. N. pp.D. J. Likewise. Use of such a method would result in the selection of four geographic isolates.I.. metabolism and detoxi®cation eciencies in two Lemnaceae. Norcross.W. Mueller.H. daylength. Evaluation of duckweed (Lemna perpusilla and Spirodela polyrrhiza) as feed for Nile Tilapia (Oreochromis niloticus).

20 B. J. Technol.L. J.D.. J.E. S.. Watt. In: Deutscher. Schenk. Cabrera. (Ed.. Phosphorus removal from static e‚uent using duckweed.V. Williams. Qual. Gholson. Ruso€. 50.L. W. Achacoso.S. A.. 1972.H. 1994. Blakeney. Porath. M. T. Bergmann et al.. 1990. 1975. 1± 15.. T. J. J. Fish Biol.. Van. 53. E.P. D.E. Oron. Culley Jr. 43±49. Guide to Protein Puri®cation.L.. D. 199±204. M..V.. Dairy Sci. / Bioresource Technology 73 (2000) 13±20 Shireman. A.L. 12 (5).T. C. de-Vegt. D.A.. Anim. Heavy metals in the environment: Bioavailability of copper.V. 179±184. 1997.D. V. Quantitation of protein. pp. R. Duckweed ± Potential feedstu€ for cattle. 23.. Bot. Dairy Sci. W.R. M. Can. Apparent digestibility of some aquatic macrophytes in rohu... Culley. 1980.. Hildebrandt.. L. D. 72±81. Aqua. I.. 28. Comparative variables of digestion and N metabolism in large white cross piglets having free access to sugar cane juice and duckweed. vol. Wood. 1972. Ray. J. Water Res. Stoscheck. 1978. 61 (s1).. A. Grith. Feed Sci.H. 367±370. New York.. Qual.H. Men. Medium and techniques for induction and growth of monocotyledonous and dicotyledonous plant cell cultures. Preston. L. Animal production and European pollution problems.. 8±9. . J....M. G. Duckweeds (Lemnaceae family): A potential source of protein and amino acids. J.). Trop. Ruso€. J.. 182. M. Duckweed (Lemna spp) as protein supplement in an ensiled cassava root diet for fattening pigs.D. Rodriguez. Sumner. J... 1996.P.. L. R. Ornes. 1994. ®ngerlings. 4 (3).R.L. D.). Zeringue. 181. Agric. Duckweed for chick feed. Sutton.W. Labeo rohita (Ham.C. H. Food Chem.. 1978. Louisiana Agric..U.. Ruso€.K. Academic Press. 9. Das. 1977. Preston. 848±850. Truax. Culley. 1988. 61 (s1)..M. manganese. Environ. Johnson. 135±144. 457±463.. Williams. 16 (1). Livestock Research for Rural Development 8 (1). P. A.H.. D. 335±342. Feed value of duckweed (an aquatic plant: Family Lemnaceae) for ruminants. and zinc in poultry litter. Son.. 1995... 50±68. 161. Gantt. L. Methods in Enzymology.. Nitrogen removal and conversion by duckweed grown on waste-water. Environ. 22 (2). Livestock Research for Rural Development 9 (1). D. Growth of grass carp fed natural and prepared diets under intensive culture. J.. B. L.