Processing, chemical composition and nutritive value of aquatic weeds


Material Information

Processing, chemical composition and nutritive value of aquatic weeds
Series Title:
Florida Water Resources Research Center Publication Number 25
Physical Description:
Bagnall, L. O.
Shirley, R. L.
Hentges, J. F.
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:


As an alternative to chemical control,water hyacinth (Eichhornia crassipes) and hydrilla (Hydrilla verticillata) can be converted to agriculturally useful products. Whole or chapped plants can be readily composted to create an organic material for potting plants. The 90 to 95% moisture content plants can be pressed to remove 75% of the water with a modest energy input. The pressed plants can be ensiled with suitable additives or dried to make animal feed. Projected best processing costs are $2.50 per ton of hyacinth silage, $11.21 per ton of dried hyacinth and $8.40 per ton of dried hydrilla. Protein contents range from 12 to 18% and ash-free crude fiber contents range from 25 to 35%, which are in the ranges usually found in land forages. Ash concentrations of 10 to 30% were found, which are much higher than those found·in land forages. Nitrates, oxalates and cyanide found in the aquatic plants were in ranges considered to be safe in usual land forages. Calcium to phosphorous ratios were at the·high end of the satisfactory range and beyond. Dried pelleted water hyacinth has a replacement value equivalent to cotton seed hulls and sugar cane bagasse. Animal acceptability of properly made hyacinth silage was very good. Animal utilization of protein was poor and that of other nutrients was fair for both dried and ensiled water hyacinth.

Record Information

Source Institution:
University of Florida Institutional Repository
Holding Location:
University of Florida
Rights Management:
All rights reserved by the source institution and holding location.
System ID:

This item is only available as the following downloads:

Full Text


Publication No. 25 Processing, Chemical Composition and Nutritive Valup of Aquatic Weeds By L. O. Bagnall, R.L. Shirley and J.F. Hentges University of Florida Gainesville, Florida es Research I


" 4 ay PROCESSING, CHEMICAL COMPOSITION AND NUTRITIVE VALUE OF AQUATIC WEEDS by L.O. BAGNALL R.L. SHI RLEY and J.F. HENTGES PUBLICATION NO. 25 FLORIDA WATER RESOURCES RESEARCH CENTER RESEARCH PROJECT TECHNICAL COMPLETION REPORT OWRR Project Number A-017-FLA Annual Allotment Agreement Numbers 14-31-0001-3209 14-31-0001-3509 14-31-0001-3809 Report Submitted: November 16, 1973 The work upon which this report is based was supported in part by funds provided by the United States Department of the Interior, Office of Water Resources Research as Authorized under the Water Resources Research Act of 1964 J


TABLE OF CONTENTS Page ABSTRACT .......... '. .. i-CHAPTER I .... : . '1 CHAPTER II ........ ......................... _.40 CHAPTER III ........ 4 ._. 45 ACKNOWLEDGEMENT 0 0 0 0 0 0 0 0 0 0 0 0 .052 REFERENCES CITED 0 0 0 0 0 0 0 0 0 0 0 0 0 053 7 1 21 I!


41 ABSTRACT As an alternative to chemical contr-01,water hyacinth (Eichhornia crassipes) and hydri11a (Hydri11a vertici11ata) can be converted to agriculturally useful products. Whole or chapped plants can be readily composted to create an organic material for potting plants. The 90 to 95% moisture content plants can be press ed to remove 75% of the water with a modest energy input. The pressed plants can be ensiled with suitable additives or dried to make animal feed. .proj ected best processing costs are $2.50 per ton of hyacinth silage, $11.21 per ton of dried hyacinth and $8.40 per ton of dried hydri11a. Protein contents range from 12 to 18% and. ash-free crude fiber contents range from 25 to 35%, which are in the ranges usually found in land forages. Ash concentrations of 10 to 30% were found, which are much higher than those found in land forages. Nitrates, oxa1ates and cyanide found in the aquatic plants were in rangeseconsidered to be safe in usual land forages. Calcium to phosphorous ratios were at the high end of the satisfactory range and beyond. Dried pe1leted water hyacinth has a replacement value equivalent to cotton seed hulls and sugar cane bagasse. Animal acceptability of properly made hyacinth silage was very good. utilization of pro tein was poor and that of other nutrients was fair for both dried and ensiled water hyacinth. i.


CHAPTER I by L.O. BAGNALL Disposal of the harvested aquatic plants was one of the most serious problems with previously attempted mechanical control systems. Disposal of reduced plant material in the water degraded water quality in the same way that present chemical control systems do. ,Disposal on the'shore produced a pile which interfered with further removal and adversely affected surrounding property values. The purpose of this project was to explore uses for the harvested aquatic plants, particularly for animal feed, and to develop processes for converting the raw plants to the most' useful products by the most economical The aquatic plants which were examined in the processing phase of the study were water hyacinth and hydrilla. Water hyacinth, shown in Figure 1 is a large, free-floating plant, thought to be easy to Stand densities range from 50 to 200 tons per acre and moisture content is typically 95 percent. Hydrilla, shown in Figure 2, is a bottom-rooted submersed plant, requiring specialized equipment to harvest in any quahtity. Stand densities range from -1 to 3 tons per acre and moisture content is about'90 percent. Further description is given by Weldon, Blackburn, and Harrison (1969), and scope of the infestation'problems are described by Holm, Weldon, and Blackburn (1969). The system to process aquatic weeds is shown in Figure 3, with products availabl'e at the completion of various stages. Though harvesting may be only marginally a processing operation, harvesting rate and condition of harvested plants affect subsequent operations. If sufficient time and spac? are available, a composted product can be made with no further processing. Chopping or reduction increases the bulk density and improves the handling characteristics of, the plants, which helps processing, transportation, and'storage. Chopping also increases biological activity in composting and reduces the time required to produce the composted product. Hyacinth have been chopped for paper-making to facilitate feeding of the attrition mill. Pressing, or fractionation, separates the ,plant into a dryer, fibrous fraction and a nutritious liquid fraction. The fibrous fraction can be dried or mixed with carbohydrate additives and ensiled. Separating the suspended nutrients from the press liquor produces a waste liquor which is less damaging to the environment and a cake which is high in useful nutrients and low in fiber. Drying produces a feed that is easier to store and transport, being lighter and less biologically active. The dried aquatic plant material mixes readily with other dietary ingredients to become part of a complete balanced rati9n. Pelleting 1


I. Water hya('iIlLh (F:iihhornia nassipt's) 2. Ilvdrilla (lI,drilla v('J'(i('illala)




increases the density, reduces the dust losses, and improves the palatability of either the pure plant feed or the mixed feed. Most of the harvesters observed or used in connection with this project were flat wire belt conveyors. The most complex was an Aquamarine harvester, a self-propelled barge with cut-ting and carrying capability, which was used to harvest most of the hydrilla. The simplest appeared to be an adaptation oE a heavy farm conveyor. Between these was the Leach hyacinth harvester, a shore-based, 10-foot wide unit, used to harvest hyacinth for the largest tests. An IS-inch bale elevator, modified by the addition of flights, sides, and deck was too narrow for harvesting hyacinth. A light 42-inch chain-and-flight conveyor, shown in Figure 4, was satisfactory when hand fed but could have been improved by perforating the deck at the pickup point, shielding the chain return loop, reversing the hitch and changing the flight configuration. When operated in excess of 50 feet per minute, pickup turbulence de-flected plants. Exploratory research on crimper harvesting of hyacinth was begun on this project. The research has been expanded and is being completed under a grant from the Florida Department of Natural Resources. The crimper harvester, shown in Figure 5, lifts, throws and partially reduces hyacinth with one simple rrlechanism. In its present state of development it is not as effective at any of its functions as individual, specialized machines. A waste handling pump, shown in Figure 6, was tested as a hyacinth harvester. At 810 pump rpm, well below maximum rated, it removed 3 tonsor 0.03 acre of small hyacinth per hour and required 3.4 horsepower hours per con, ten times as much as a conveyor. The pump chopped the plants quite satisfactorily but separation of the fragments from the water stream was difficult; this difficulty prc""r;nted complete evaluation at higher speed, where it appeared to draw in plants more aggressively. The pump was mounted at about a 300 angle from the horizontal and operated near the surface so that an open vortex formed to ingest plants; it did not work at all when operated vertically or deeply. Modified forage harvesters were tested and used for production. The shear-bar chopper produced relatively uniform material in lengths of 3/8 to 1 inch. The International 350 harvester, shown in Figure 7, had a l6-inch wide chopping cylinder and -was equipped with 3 blades and operated at 1000 rpm to produce oneinch cuts. It chopped 28 tons per hour (rated capacity: 40 tons of corn per hour) and required 0.35 horsepower hours per ton. Specific energy was proportional to the square of the cutterhead speed and was little affected by feed rate, length of cut, shear-bar clearance, and minor changes in feed geometry. When the scroll was removed and the discharge directed downward 4


Figure 4. Hyacinth harvesting conveyer Figure 5. Experimental hyacinth crimper-harvester


6. Figure 7. 'Ii. .. ::." <:\ -.-", WUlite pump Lest ali a Forage harvesler ehoppillg water hyacinth ...... "\>:


.. rather than upward the spout power requirement increased 50 percent because of excess.ive recirculation. The standard feed mechanism was inadequate for hyacinth. Confining the feed stream improved capacity and specific energy requirement to that given above. A more aggressive feed mechanism would certainly have increased capacity and as result would probably have decreased specific energy. Because the spout plugs frequently when chopping fresh hyacinth with the cut-and-throw chopper, a simple cutter--withchain-and-flight conveyance is recommended. Bulk density of chopped hyacinth is 29 to 53 pounds per cubic foot, depending on degree of packing, as compared to 5 to 6 pounds per cubic foot for Whole plants. 'A crimper, shown in Figure 8, was devised to reduce hyacinth. The rolls are 6-inch pipe with 1/4 x I teeth and operated at 800 rpm. The 3 horsepower engine. initially used was replaced with a 5 horsepower engine because slugs of hyacinth stalled the machine. Pneumatic cylinders were used initially to load the floating roller but were replaced with springs after a roller. loading .of 200 pounds per foot was found to be adequate; this allowed field operation of the machine without auxiliary tanks or compressor. When operated at the end of the conveyor in Figure 4, capacity was two tons per hour. Because the plant stream converged, slugging was aggravated; this type of device should be designed to the full width of the harvester conveyor with which it is used. Reduction of the plants is not as good as that of the shear bar chopper but cost is lower. Leach's commercial harvester uses a flail chopper over the cross conveyor to reduce fresh hyacinths. Capacity is high and bulk density and biological activity are increased satisfactorily. The product is non-uniform and in some cases does not feed satisfactorily to subsequent processing. When whole or chopped hyacinth are piled and allowed to decompose aerobically, a is formed. Whole plants require about six-months u.l!"';' chopped plants three months to compost adequately for commercial use. The compost is'then dried, ground, and mixed with mineral constituents. This material has been used for potting ornamentals and as an additive to municipal park flower beds; it has sold for $12 per cubic yard. Preliminary tests show the hyacinth compost to have excellent water retention and indications are that it can constitute no more than 25 percent of a sand-compost mix without harm to the plant. Tensile, shear, and compressive strengths of fresh mature water hyacinth stems were found by loading individual stem samples in an Instron tester. Ultimate strengths were 109 psi in axial tension, 28 psi in axial compression and 43 psi in lateral double shear. Buckling failure, of the shortest compression samples and jaw compression of'the tensile samples indicated that the pith carried little of the load. Ultimate tensile strength of the loadbearing fibers was 886 psi. 7


Figure 8. Hyacinth crimper


Small samples of chopped and minced water hyacinth were pressed in 2 1/2 to 5 '1/2-inch diameter cylinders lined with perforated stainless steel. Expression increased with increased pre-grinding, increased pressure, decreased diameter, increased liner open area, increased pressure rise time, and increased number of pressings. There was little additional expression at pressures above 100 psi or at rise times longer than 40 seconds. Pressure and specific volume were related by v = 1.303 p-.07496 (r = .99) where V = specific volume, glcc P = pressure, psi r = correlation coefficient. Expression energy found in these tests is shown as a fUnction of percentage expression in Figure 9. Energy L:'-lnd in the static tests is much lower than that of production machinery. Three commercial screw presses were used in early tests and some aspects of performance measured. One was a machime of unknown manufacture used on the processor, shown in Figure 10, provided to the Florida Game and Fresh Water Fish Commission by Stanley Hiller; the others were six-inch and twelve-inch Vincent screw presses, shown in Figuresll and 12, previously used for citrus pulp and land forages. Configuration of the first press, shown in Figure 13, featured a tapered shaft,'uniform pitch with the exception of the feed section, aligned cutting edges, bevelled teeth, and an adjustable, fixed-clearance, rotary discharge cone. The press consumed 2500 pounds of hyacinth per hour and expressed 75 percent of the water and 15 percent of the dry matter, including 21 percent of the protein, at a cost of about 16 horsepower hours per ton of water expressed; peak power equipment was much higher. The press consumed 3600 pounds of hydrilla per hour and expressed 80 percent of the water and 37 percent of the dry matter, including 34 percent of the protein. Press power requirement was limiting factor in production of the processor. Configuration .of the Vincent press, shown for the twelveinch press in Figure 14, featured a straight shaft except near the discharge, decreasing pitch, non-aligned cutting edges, relieved teeth, and a pressure-backed, floating, non-rotating discharge cone. The six-inch press consumed 216 pounds of hyacinth per hour and expressed 80 percent of the water and 23 percent of the dry matter including 36 percent of the protein at a cost of 18 horsepower hours per ton of water expreseed; it consumed 40 pounds of hydrilla per hour and expressed 85 percent of the water and 32 percent of the dry matter, including 41 percent of the protein. Production was limited by components in the system fore the press. The twelve-inch press expressed 74 percent of the water and 32 percent of the dry matter, including 34 percent of 9


I-' 0 ........ 0.16 0 w Vi Vi W 0:: CL X W .0:: 0.121 TWO PRESSINGS", W lL. 0 o.OSJ ONE PRESSING '" I-...... Vi 0:: ::> 0 I I 0:: w 0.04 3: 0 CL W Vi 0:: 0 I G 0 0:: W 0 20 40 60 80 100 z w PERCENT OF INITIAL WATER EXPRESSED FrGURE 9. WATER HYACINTH WATER STATfC EXPRESSION ENERGY ----------


Figure 10. .. Mobile aquatic plant processor ,.. Figure 11. Six-inch Vincent screw press with casing removed


Figure 12. Twelve-inch mobile Vincent screw press


13 W -I CO o 2 z o tr 0::: => (9 LLo::: Zo OU) UrJ) If)W If)U wO 0:::0::: CLCL (Y) 't"'""" ....... _I I


If) If) W Ct: D-s W Ct: U If) W .:..] dJ o 2 I Uz Zo Il(\j or-Ct: l-.=> z<.9 WLL UZ Zo >U


the protein, from 25,000 pounds of hyacinth per houri it removed 87 percent of the water and 43 percent of the dry matter, including 60 percent of the protein, from 3400 pounds of whole hyacinth per hour. Estimated power requirement at the high production rate was 4 horsepower hours per ton of water expressed or 2.5 horsepower hours per ton of chopped hyacinth. A family of lightweight screw presses was developed, including two eight-inch designs, two nine-inch designs, and one twelve-inch design. features ,are: (a) extra heavy constant pitch helicoid conveyor screw with pitch equal to length, modified by notching, (b) fixed teeth projecting from the casing halfway to the shaft, (c) lightweight circumferentially ribbed casing line with 30%-open 16-gauge perforated steel sheet, (d) 900 included angle conical discharge restriction, supported by screw shaft, non-rotating and axially constrained, and (e) screw speed near 50 rpm. The first eighi-inch press, shown in Figures 15 and 16, showed that a continuous flighted screw operating in a casing with flame-cut drainage slots and no teeth would not satisfactorily dewa-ter hyacinth at an adequate rate The second eight-inch press, shown in Figures 17 and 18, performed adequately with two pairs of teeth and a 26 1/4-inch long screw, double-flighted in the pressure sections. It expressed up to 71 percent of the water and 25 percent of the dry matter from up to 1.7 tons of chopped hyacinth per hour. Expression was dependent on discharge restriction pressure. The screw and housing are segmented so that variations of configuration, primarily pitch and length, can be readily examined. The first press in the family, the nine-inch press shown in Figures 19 and 20, was completely self-contained and was used as a portable small-scale production machine. It had three sets of teeth and a 31 1/2 inch long single-flighted screw. The casing was a welded fabrication of 1/4-inch wide by I-inch high bars. It expressed up to 65 percent of the water from hyacinth in one pressing and up to 77 percent in two pressings a-t 22 psi projected restriction pressure; expression was pressure dependent. It pressed up to 3.8 tons of chopped hyacinth or 4.2 tons of prepressed hyacinth per hour and consumed an estimated 4 horsepower hours per ton of water expressed or 1.9 horsepower hours per ton of chopped hyacinth. Single flighting in the screw pressure sections produced an unbalanced load which caused the cantilevered screw to gyrate; in succeeding models, the pressure sections were double flighted. High labor cost and warpage of the fabricated casing led to development of the pipe casing used in subsequent models. The press was released to USAID and sent to Bangladesh to assist in a development program there. The second nine-inch press, shown in Figures 21 and 22, incorporated devised as a result of observation of 15


Figure 15. Simple eight-inch press-disassembled (SSA)


17 If) If) w-0:: D--3: w 0::-U If) I Z uO z r--' roO:: ::) wl9 --l-LL D-z-2 --_0 If) -u to


_ Figure 17. Variable configuration B-inch press in initial configuration (SBB) "-_fN'fiSEZKFW TM




Figure 19. First 9inch press, partially disassembled to show cage hoops (S9A) -'"---------.;..----------------........;.---------


:.-=.==== 21 W O:::z Uo If)-I 0.::. U:) 2(.9 ILL Wz ZO zU o C\J W 0::: <.9 LL


Figure 21. 1m pro ved 9 inch press (S9C)


, r"--+-1 "I 23 U (J) If) '-.-/ if) If) W 0::: CL W 0::: U If) I Uz 00:: W:) ><..9 OLL 0:: Z CLO 2:u C\l C\l w n:::: :) <.9 LL


its predecessor. The screw was the same length and was segmented at the same points, but was double-flighted in the pressure sections to reduce gyration. The casing was flame-cut from a 3/8-inch wall pipe and cantilever-mounted. The space-frame was eliminated and the drive simplified to be ammendable to gasoline engine, electric motor, hydraulic motor or PTO; weight was reduced 20 percent. Preliminary tests indicate that production rate is similar but that other performance para..Yfieters are improved by the design changes. An identical unit is being used in a test program in Guyana. The twelve-inch press, shown in Figures 23 and 24, was similar in configuration to the second nine-inch press, having three pairs of teeth and a 43 7/8-inch long screw. Originally set up with two parallel screws as a 25 ton-per-hour machine, it was reduced to one screw when feed and discharge interference were found to reduce production to that of one screw. It is PTO driven and has been used as a large scale production unit, often with direct feed from a harvester conveyor. Performance has not been established and it is being modified prior to performance testing. In summary, the lightweight presses designed in the course of this project were adequate in strength and durability during the test period. Performance of the commercial design was better and would probably be more economical in the long run. The most satisfactory design may combine the light of the project presses with the sophistication of the commercial screw configuration. Chopped and pressed water hyacinth were ensiled in barrel, culvert, and tower silos for periods ranging from 25 to 217 days. Runoff quantity and pH, weight loss, shrinkage, and spoilage were measured and quality of product was observed. Chopped hyacinth shrank and drained excessively and putrefied in the silo. Pressed hyacinth silage was marginally acceptable. Pressed hyacinth ensiled with two to four percent dried citrus pulp or cracked yellow dent corn as a carbohydrate source drained, shrank and spoiled little and in most cases was very acceptable. Drainage in the barrel silos decreased with decreasing moisture content and stopped altogether below 85 percent moisture. When the silage was properly packed and drainage was low, spoilage, shrinkage, and weight loss were also low. Molasses at low levels (0.5% to 1.0%) did not improve silage quality. Analysis of the juice showed that the chemical oxygen demand (12.000 mg/l) and nitrogen content (650 ppm) were too high to allow its disposal by introduction into public waters. Up to 63 percent of the dry matter in the juice was recovered by filtration and low-speed centrifugation. Filters fine enough to recover more than 10 percent of the dry matter blind easily and very little liquor per unit area per unit time can be processed, so filters must be designed carefully to avoid blinding. Filter cake is fairly high in moisture content and fiber and could be 24


Figure 23. Twelve-int.:h mobile, PTO-driven press (SI2A)


26 tJ) (j) w n:: (L 2 (L W -1 (]) o 26 I ul--. W(9 wz C\J W 0::: <.9 LL


dewatered prior to drying. Centrifugation produces a product higher in xanthophyll and protein than drying or filtration. Juice quantity and characteristics vary with plant characteristics and processing and a universally applicable system must be tolerant of this variability. Increasing centrifuge speed increases dry matter recovery and precipitate dry matter content and decreases supernatant fluid dry matter content. Protein and xanthophyll contents are oisappointingly low; similarly prepared leaf protein products usually show much higher levels. A student working on a special problem examined some of the drying characteristics of crushed water hyacinth in a 6-inch diameter static column at 1500 cfm per square foot flow rate. He found the drying equation to be where M -M 2 ___ --=e= e-O. 25 DLt 1f 2/d M M o e M = moisture content, dry basis, at time t Me = equilibrium moisture content, dry basis M = initial" moisture content, dry basis o DL = diffusivety, ft2/hr -t = time, hours d = bed depth, feet. He further found D = 10-54. 2 T19. 5 L where T = temperature, oR. Drying rate was strongly dependent on moisture diffusion within the material. and was not much affected by flow rate and relative humidity within the range tested. Equilibrium moisture content was related to temperature by M 43.6 -.0637T (correlation significant ata=.05). e In drying tests with 15 to 50 pound samples of hyacinth and hydrilla it was found that agitation, flow rate, heat rate, preprocessing, initial weight, and plant species affected drying rate and that size of whole hyacinth did not. The drying data was regressed to where M -M e = e -kt -1 k = drying constant, hours The drying constants for the various treatments are shown in Table 27


------------------------_._-_._-----. 1. All reported effects are significant (a = .05) and most are highly significant (a = Ol). Tumbled samples dried five times as fast as the most nearly comparable static samples. Large samples of hyacinth dried faster at high flow rate, but small samples showed a mixed response; drying rate of small hydrilla samples decreased with increasing flow rate. Tripling heat input rate almost tripled drying rate. Crushed hyacinth dried 1.5 times as-fast-aswhole plants, pressed hyacinth and finely shredded hyacinth dried 1.8 times 'as fast, and coarsely shredded hyacinth dried 2.4 times as fast; finely shredded hydrilla dried 2.1 times ,as fast. as whole hydrilla plants and crushed hydrilla dried 2.4 times as fast. Drying rate decreases with increasing sample size. Hydrilla dried twice as fast as hyacinth. A 6000 pound-per-hour triple-pass rotary forage dehydrator was used to dry ten tons of pressed hyacinth. As operated, it evaporated 5800 pounds of water per hour while reducing the hyacinth from 88 percent to.22 percent moisture content; energy required was 1500 BTU per pound of water evaporated. The product was granular and relatively free-flowing. It was not quite dry enough for safe storage, but could have been made so with minor modifications of operating procedures. Lots of 3000 pounds of dried hyacinth and hydrilla were a pelleted at the Coastal Plain Experiment Station, Georgia The feed was hammermilled and steamed prior to pelleting. Pro duction rate was low due to the poor flow characteristics of the material in the feeder and in the die, and power required per ton was unreasonably high. A very dense (bulk density = 52 p01:lnds per cubic foot), durable (durability index = 98) pellet was formed. Complete diets using hyacinth and hydrilla with corn and soybean meal were pelleted satisfactorily the same mill at 2 tons per hour and 19 horsepower hours per ton. -Pellet density (bulk density = 43 pounds per cubic foot) and durability (durability index = 75) were lower. Differences in pelleting characteristics were due to lubricity and feeding characteristics of the other feed constituents and to regrinding of the aquatics. Four processing systems, fragments of the overall system, were used to produce feed products and major aspects of their performance recorded and analyzed. Hyacinth and hydrilla were processed in the Hiller processor, shown in Figure 10 and, schematically, in Figure 25. Hyacinth and hydrilla were processed in a citrus pulp dehydrating pilot plant, shown schematically in Figure 26. Hyacinth was processed with the mobile Vincent press to produce silage, and with the mobile press and a stationary dehydrator as shown in Figure 27 to produce a dry, granular feed. acooperation of J. L. Butler and R. G. Hellwig, ARS-USDA, was' essential to. this phase of the program and is appreciated. 28


., Table 1. 'Aquatic Weed Drying Constants, hours1 Heat Ineut (watts} 500 1000 1500 Other Hyacinth Whole, 20 pourids, 70 cfm .0416 .1006 .1632 a .1064 b .1590 Coarse shredded, 15 pounds, 70 cfm .1357 .2691 .42)5 20 pounds, 35 cfm .1526 45 cfm .1492 70 cfm .2546 110 cfm .2449 50 pounds, 45 cfm .0507 70 cfm 90 cfm .0862 Fine shredded, 20 pounds 1.3001 c 35cfm .1077 70 cfm .0765 .1467 .2465 ". Crushed, 20 pounds, 35 cfm .0973 70 cfm .0965 .1209 .1832 Pressed, 20 pounds 1.8125 c 35 cfm .1273 70 cfm .0942 .1119 .4130 Hydri 11 a Whole, 20 pounds 35 cfm .1187 70 cfm .0904 .1375 .2229 Fine shredded, 20 pounds 35 cfm .3139 70 cfm .1757 .3146 .3156 Crushed, 20 pounds 45 cfm .3500 70 cfm .3289 110 cfm .2823' a 140 F constant temperature b 180 F constant temperature c Clothes tumbler dryer, high temperature, unknown flow. 29








The Hiller processora included the chopping, pressing, and drying operations; complete process, as used included harvesting, additional drying, and pelleting and produced pelleted cattle feed as its end product. Juice was wasted. Hyacinth was harvested by raking the floating plants onto a light industrial conveyor which lifted them from the water to the beed hopper. Hydrilla was harvested by an Aquamarine harvester piled on the bank and fed with the same elevator. The hyacinth was chopped by counter-rotating lawn-mower reels; thehydrilla was not chopped. Press operation and performance have been described. The dryer was a tube, with agitators, through which the pressed residue was carried by hot air and combustion products. It dried hyacinth residue from 85 percent to 73 percent moisture content, removing 63 percent of the water, at an expenditure of 1800 BTU per pound of water evaporated; it dried hydrilla residue from 71 percent to 63 percent moisture content, removing 50 percent of the water, at an expenditure of 2100 BTU per pound of water evaporated. Both processes are quite inefficient, especially considering the high final moisture levels. The products were too wet for safe storage and were dried in static-bed gas-fired cabinet dryers at about 1400F. Some deterioration and molding took place during finish drying. The pelleting operation has been described. Estimated production costs with this system are shown in Table 2. Low capacity and efficiency of this system make it uneconomica,l. The citrus pulp pilot plant consisted of a shredder (equivalent to a chopper), a press, and a 600 pound-per-hour triple-pass rotary dehydrator; the complete system also included hand harvesting. Test runs consisted of about 400 pounds each of hyacinth and hydrilla. Shredding energy requirement was high. The slow speed screw conveyors plugged easily. During the hyacinth test the press cage liner burst. The dryer was inefficient because of low loading and did not dry adequately, partly because of the 10v1 head temperature (500-7000F). It is likely that many of these deficiences could be overcome with experience in a larger scale plant but shredding and screw conveying of wet aquatics would have to be avoided. Production was so low that reasonable economic projections could not be made. The mobile press silage system consisted of a harvesterchopper and the mobile Vincent press. This system was used to place 44 tons of silage in a tower silo. The press was the capacity limiting device in system and this limitation was aggravated by the coarseness and variability of the flail-chopped aThe cooperation of Vernon Myers and his crew from the Hyacinth Control Division of the Florida Game and Fresh Water Fish Commission in providing and operating this equipment is appreciated. bThe cooperation of John Moore of Sweeney Environmental Controls in providing the harvester operator for this operation is appreciated. 33


Table 2. Estimated Harvesting-Processing Cost with Mobile Processor MACHINE ENERGY LABOR 34 $/raw ton 10.13 _0.98 5.62 16.73 $/dry ton 238.29 23.05 -132.35 393.35


material. Estimated silage cost is given in Table 3. The cost is high primarily because of low press capacity The mobile press-stationary dehydrator system included a harvestera chopper, press and dryerb Operation and performance of all elements of this system have been described previously. Estimated operating costs are given in Table 4. It will be noted that all estimated operating costs put the cost of the product unreasonably high. This is partly because operation is assumed to be on a small scale, approximately the experimental level. Conditions were those observed and not necessarily the best observed for the same equipment. There were some discrepancies in capacities of individual devices and in every case one device limited the performance of the system. Material balances of optimized hyacinth and hydrilla processing systems are shown in Figures 28 and based on 100 pounds of dried product of typical analysis. Estimated production costs in these systems are shown in Table 5, based on continuous operation of 1000 acres of hyacinth or 7000 acres of hydrilla. These costs do not include harvesting, materials handling marketing, and so are low; but they are comparable to the figures in the previous tables. aThe cooperation of T.W. Casselman and other personnel of the Agricultural Research and Education Center, Belle Glade, in providing equipment, personnel and skills is appreciated. bThe cooperation of C.D. Leach and his crew from Sarasota Weed and Feed in providing and operating the harvester is appreciated. 35


Table 30 Estimated Cost of Producing Silage $/raw ton $/dry ton MACHINE 3.09 61.73 ENERGY .12 2.33 LABOR 3.49 69.77 ADDITIVES 1.00 14.40 7.70 148.23 Table 4. Estimated Harvesting-Processing Costs of Mobile PressStationary Dehydrator System. $/raw ton $/dry ton MACHINE 2.71 85.86 ENERGY 1.04 32.86 LABOR 1.17 37.09 4.92 155.81 36


2381 chopped hyacinth 2262 water 450 100 1715 119 solids 17 protein 20 ash _----------9700 BTU pressed hyaci nth 360 water 90 solids 10 protein 10 ash ____ BTU water dried hyacinth 10 water 90 solids 10 protein 10 ash solution 1708 water 7 solids 1 protein 5 ash 1931 216 juice 1902 water 29 solids 7 protein 10 ash residue 194 water 22 solids 6 5 ash 288,000 BTU ]92 water Figure 28. Hyacinth Process Material Balance 37 24 dried residue 2 water 22 solids 6 protein 5 ash.


1944 chopped hydri11a 1788 water 156 solids 19 protein 42 ash _-------8000 BTU 299 pressed hydri11a 209 water 90 solids 11 protein 21 ash 298,000 BTU >----------:11 ..... 199 water 100 dried hydri11a 10 water 1154 90 solids 11 protein 21 ash solution 1237 water 17 'solids 2 protein 10 ash 1645 juice 1579 water 66 solids 8 protein -21 ash 491 residue 442 water solids 6 protein 11 ash 656,000 437 water Figure 29. Hydri11a Process Material Balance. 38 54 dried residue 5 water 49 solids 6 protein 11 ash


Table 5. Estimated Processing Costs Under Optimum Conditions Hyacinth Silage $/raw ton $/silage ton a $/dry ton Machine 0.036 .11 0.73 Energy 0.061 .18 1.21 Labor 0.018 .05 0.35 Additives 1.000 2.16 14.40 1.115 2.50 16.69 Dried Hyacinth $/raw ton $/product ton b $/dry ton Machine 0.086 1.84 2.04 Energy 0.385 8.80 9.78 Labor 0.032 0.65 0.73 0.503 11.29 12.55 Dried Hydrilla $/raw ton $/product b ton $/dry ton Machine 0.077 1.20 1.33 Energy 0.287 5.06 5.62 Labor 0.146 2.22 2.47 0.510 8.48 9.42 a 15% dry matter b 90% dry matter 39


CHAPTER IT by R.L. SHIRLEY Studies have been pursued in the evaluation of (a) potential toxicants, (b) nutrient element and sUbstance composition, and (c) performance of rats fed diets containing aquatic plants. Toxicants Nitrates: Hyacinths contained approximately 0.3% nitrate during April that were obtained from Lake Apopka, the Kissimmee River, and the St. Johns River; but, generally throughout the year of monthly samplings values ranged 0.05 to 0.1%. Apparently nitrate concentrations above 0.5% are required before any observed harm to livestock occurs. In digestion trials with steers, hyacinths, hydrilla and coastal bermudagrass rations varied from 120/ 9/ 5 mgNo3 per kg, respectively; and the digestibility or destruction of Ehe dietary intake was 88, 75 and 83%, respectively. Oxalates: Concentrations in hyacinth over a year of monthly sampling varied from approximately 0.2 to 0.6% (avg. 0.4%). If a steer consumed 14 kg dry weight of ration containing 0.4% oxalate, the oxalate could bind 25 g of calcium as insoluble oxalate. However, in digestion trials with steers 80, 78 and 91% of the oxalate in hyacinth, hydrilla and coastal bermudagrass rations failed to appear in the feces and urine, respectively. Apparently the oxalate was reduced to carbon dioxide and water in the rumen. Cyanide: There were a few analyses when cyanide approached concentrations of 30 LUg per kg fresh weight. Sorghums have been found to be safe to graze if they contain less than this amount of cyanide. Pesticides: Hyacinths were collected at monthly intervals over apprOXimately a one-year period from several lakes and rivers and portions submitted for pestiCide analyses to W.G. Fong and D.C. Golden, DeparDuent of Agriculture, Tallahassee, Florida. Hyacinths from Lake Santa Fe were free of pesticides from March through December. Bivans Arm Lake hyacinths had traces of DDD and DDE in May and a trace of diazinon in December. Those from Lake Apopka had traces of DDT in March, June and September; DDE traces in March, June/ July, August, October and November; and DDD traces in June through November. KissimmeeRiver hyacinths had a trace of DDT in March; and traces of DDT and DDD in May. The St. Johns River hyacinths had traces of DDT, DDD, and DDE in Marchi traces of DDT and DDE and 0.18 ppm of DDD in May; and a trace of diazinon in December. Organic Nutrient Composition of Aquatic Plants (Dry Weight Basis) Protein: Generally 12 to 18% protein was present. Hyacinths 40


grown in lakes were more variable than those grown in rivers as observed over a year of monthly sampling. Ether extract: Usually Ito 2.5% was present and quite similar in concentration to land forages. Crude fiber: Generally 13 'to 20%. If expressed on the ashfree dry weight basis it would have been somewhat higher and more .equivalent-_ to land forage values-of 25 to 30%. Xanthophyll: This pigment ranged from approximately 330 to 550 mg/kg of dry hyacinth in Lake Apopka during spring months; corresponding values for other lakes and rivers ranged from 88 to 250 mg/kg over a year of monthly sampling. These values compare favorably with reported values of 170 mg/kg of 17% protein alfalfa. Carotene: Values ranged in hyacinth obtained in Lake Apopka from 66 to 77 mg/kg in March and May; and corresponding values in other lakes and rive-rsvaried from 18 to 46 mg/kg over a year of monthly sampling. Inorganic Nutrients in Aquatic Plants Ash: Generally ash was the most variable component in that concentrations varied from 10 -to 30% or above, compared to land forage values in the range of 5 to 8%. This fraction of the plant may need the most consideration in the proper utilization of aquatic plants in animal diets. Percentage of the daily National Research Council (N.R.C.) requirements of nutrient elements provided 450 kg steers per kg dry aquatic plants (average concentration): In Table 6 are presented the percentages of the daily N.R.C. requirements of Ca, P, K, and Nafor a 450 kg' steer that are present per kg of hyacinth and nydrilla (dry basis). These data show that one kg of dry hyacinth on the average provides 184, 41, 74, 84 and 134% of the steers daily needs for Ca, P, K, Mg, and Na, respectively; while corresponding percentages per kg of dry hydrilla are 670, 25, 76, 83 and 108%, respectively. These values suggest that approximately 3 kg of dry hyacinths would provide an excess of the major nutrient elements. The hydrilla on the average provides over 6 times the daily requirement of Ca per kg dry weight with a Ca:P ratio of approximately 25:1. A Ca:P ratio of 2:1 or not more than 4:1 is generally'recommended in livestock feeds. More limited data indicates that this is the situation with Ca and P and other elements with pondweed, horn wort, eelgrass and naiad. One kg of dry hydrilla was found on the average to provide 142, 59, 165 and 347% of the daily requirements of steers for Cu, Zn, Fe and Mn, respectively. Corresponding values for these trace elements in hyacinth were 15, 29, 172 and 289%, respectively. Pro cessing to remove the water may drastically alter the nutrient element composition as it does protein. Nevertheless, the high 41


Table 6. Percent Daily N.R.C. Requirements of Nutrient Minerals Provided steers (450 kg body weight) Per kg Dry Aquatic Plants (Average Concentration) N.R.C. Daily % Daily req./kg dry matter Req. s Hyacinth Hydr il1a Macroelements: gm Calcium 12 184 670 Phosphorus 12 41 25 Potassium 54 74 76 Magnesium 7 84 83 Sodium 7 134 108 Microelements: gm Copper 79 15mg 142 Zinc 150 29 59 Iron 990 172 165 Manganese 49 289 347 ----------------------------------------------------.--------------------42


concentrations of nutrient minerals in aquatic plants suggest care should be taken if they are used in livestock rations to prevent mineral imbalances. Availability of nutrient elements in aquatic plants fed to steers: In Table 7 data (Stephens, M.S. Thesis, 1972) are presented on the intake and retention of Ca, P, Mg, Na, S, K, Cu, Mn and Zn by steers in a digestibility trial when hyacinth, hydrilla and coastal bermudagrass made up approximately 33% of the organic matter of the rations. The very high level of Ca in the hydrilla diet was likely the cause of the relatively low retention of P in this ration. Mg retention was lowest in the hyacinth ration but similar concentrations and utilization occurred in the three forage diets. Na, S, K and Zn did not vary markedly in the aquatic and land forage rations, or in retention by the steers. Fe and Cu retention were relatively low with the hydrilla ration. Mn retention was relatively high with the hydrilla ration. Overall, the retention of the various nutrient elements in the aquatic plant rations were comparable to those of the land forage. Growth and qestation of r.ats fed aquatic plant diets: A study was made to compare the nutrient and possibly toxic quality of two prevalent aquatic plants to a high quality land forage for growth and reproduction in rats. Forty female weanling Sprague Dawley rats were randomly divided into three dietary groups. All diets contained 13% casein, 25% sucrose, 5% corn oil, 5% H.M.W. salts and 2% vitamin fortification mixture. To these basic constituents diets If 2 and 3 were made up by adding 50% air dry hyacinth (Eichhornia crassipes), hydrilla (Hydrilla verticillata), and alfalfa (Medicago sativa), respectively. The average weight gains during the first nine weeks on the diets were 117, 98 and 130 gm for the hyacinth, hydrilla and alfalfa dietary groups, respectively. The corresponding weight gains up to 12 weeks on the diets were 156, 142 and 170 grams, respectively. The number of young littered by the hyacinth, hydrilla and alfalfa groups were 44, 48 and 48, respectively. While none of the dietary groups grew as rapidly or had as many offspring as corresponding rats on stock rations, there were no apparent evidence of lack of vigor or health. 43


I Table 7 -Intake and Retention of Dietary Minerals as Com]2ared With the NRC Reguirements Per Day 1 Diet COASTAL BERMUDA HYDRILLA HYACINTH Intake Retent'n NRC Req Intake Retent'n NRC Req Intake Retent'.n NRC Reg Mineral Mineral Intake 2 3 Mineral Mineral Intake 2 3 Mineral Mineral Intake 2 3 Per Day Per Day Per Day I Per Day Pe:J:" __ I Day Per Dc3.-Y ___ Per Day 1 Ca, g 34.8 19.0 1-i.O 204.0 31.3 14.0 46.7 9.8 14.0 P, g 15.3 6.5 14.0 +1.6 2.9 14.0 17.6 6.1 14.0 Mg, g 12.3 3.6 5.0 18.7 3.9 5.6 15.6 1.1 5.1 Na, g 20.7 9.6 8.9 23.7 12.9 7.0 29.6 14.9 7.9 S, g 10.5 1.7 8.0 8.6 1.i 6.3 12.5 0.9 7.1 K, g 34.7 16.1 48.0 44.1 20.3 37.7 40.3 16.9 42.7 Fe, g 2.5 1.0 0.6 1.6 0.2 0.5 1.0 0.6 Cu, mg 41.3 28.3 48.2 18.0 6.0 37.7 .41.3 26.5 42.7 Mn, mg 204.5 100.5 40.2 86.5 79.1 31.5 437.0 237.3 35.6 Zn, mg 29.3 20.2 80.4 137.7 21.3 .62.9 23.0 17.0 71.2 lAverage dry matter intake by steers for diet (in kg): Coastal bermuda, 5.60; Hydril1a, 4.15; and Hyacinth, 5.19. 2Based on requirements expressed generally in mg or g per kg dry matter of ration consumed. 3calcium and phosphorus values are based on requirements for 0.5 kg weight gain per day for 300 kg steers.


CHAPTER III by J.F. HENTGES, JR. During the 88 years which have elapsed since the introduction of the water hyacinth, Eichhornia crassipes, into the U.S.A., it has regressed-from-a tropical plant prized for its colorful exotic blooms to an environmental menace. Control of E. crassipes by chemical methods'is obstructed by legal and logistical barriers. The currently popular concept of biological control is faced with a lengthly period ofexpel;imentation before safe and economical control methods are perfected. The objective of this report is to project an image of the water hyacinth as a natural resource which can be mechanically harvested and processed into a potential source of animal food. "'. "';.P Although statements have appeared in journals to indicate that E. crassipes was fed to pigs in Asia, was grazed by cattle in the tropics and was hand harvested during droughty weather as fodder for ruminants, there is a paucity of scientific literature on the nutritive value of this floating aquatic plant. Extensive studies on the nutrient composition of E. crassipes have been reported by Boyd (1968a, 1968b, 1969,.1970), Datta, et al (1966), Shirley (1970) and Taylor (1968,1969). The relationship of the chemical compositi.on of E. eras sipes to fertility of water at the site of harvest has been reported by Chadwick and Obeid (1966), Denton (1967), Lawrence and Mixon (1970) and Steward (1970). Yields per hectare of E. crassipes have been reported by Boyd (1970), Penfound and Earle (1948), Steward (1970), Yount (1964) and Army Corps of Engineers (1946). E. crassipes has been studied as a food source for fish (Liang and Lovell, 1970), swine (Combs, 1973), cattle (Hentges, 1970,1972; Salveson, 1971; Stephens, 1972; Vetter, 1972) and 1973). Classification of the numerous possible feed products derived from E. crassipes is essential for meaningful comparisons of experimental results and for commercial trade purposes. Pen found and Earle (1948) suggested the following plant size classes: midget, small, medium, large and giant; the midget being rooted on land, small being in full flower in shallow water; medium existing in still water often profusely flowered, large and giantic sizes thriving in moving, well-aerated water of canals or open expanses. The latter are distinguished by elongate, equitant leaves up to 50 inches long with float leaves being non-existent. A proposed classification by Hentges and Baldwin (1973) for E. crassipes feed products accord-ing to the international feed nomenclature (Harris, et aI, 1968) is appended to this report. 45


-\ Research on the nutritional value of aquatic plants for live,.... stock was initiated at the University of Florida in 1969. The feed products were E. crassipesand Hydrillaverticillata (Hydrilla which had been harvested from freshwater sites and processed by chopping, wet pressing and dehydration of the press residue. This crude method of processing resulted in a low quality press residue because a large portion of the nutrients were in the press juice. See Table 8. Although_research by-Taylor (L968, 1969) showed that press juice residue (cake) was rich in nutrient content, the digestibility of these nutrients by animals has never been investigated. Table 8. Losses in juice of chopped pressed E. crassipes E. crassi}2es Bydrilla sEt>-Water, % 75 80 Dry Matter (DMf, % 15 37 Ash, % of DM 50 60 Crude Protein, % of DM 15 34 Such research awaits the development of a commercially feasible method of drying the juice residue. The first question to be answered was"Nill cattle eat dried water hyacinth?". Ruminating cattle were offered diets containing processed (small, whole, chopped, press residue dried) E. crassipes as the sole diet and in mixed diets with molasses or molasses and oil seed meal. Voluntary intake did not exceed 1% of live bodyweight until the product was hammermill ground and blended with sugarcane molasses, 30% by weight. At this level intake the cattle were losing weight. By contrast similar cattle offered bermuda grass Cynodondactylon, mature hay maintained tneir weight by voluntarily consuming a quantity equal to 2% of their live bodyweight per _day. Pelleting the haromermill ground press residue increased its daily intake by cattle to about 1.5% of their live bodyweight. At this point, various levels of E. crassi}2es press residue were tested in blends with other ingredients in cattle diets. It was concluded that with the low quality experimentally processed E. crassipespress residue available at that time, an expected intake by yearling cattle of at least 2.8% of live body weight would not be attained with more than 25% water hyacinth in a balanced finishing diet. Knowing that dried E. crassipes were acceptable at a low level _in cattle diets, the next question was "Are they toxic?". In numerous short-term experiments, no signs of toxicity were observed in cattle or sheep. One group of six yearling cattle were offered dried E. crassipes at a maximum tolerance level in their diets for nine months without apparent toxic effects or digestive disturbances as judged by live performance and postmortem examination of organs and tissues. 46 $


An observation made during all of the preliminary animal feeding experiments was the need for a high level of supplemental protein in the diet. It had been learned that some protein was lost in the press juice and that Taylor (1969) found protein to be extremely difficult to extract from press residue of whole, medium size E. crassipes plants. To compare the digestibility of crude protein (N x 6.25) and dry organic matter in diets containing two aquatic plants and one standard land forage, Salveson (1971) used yearling steers in metabolism crates. The test materials were E. crassipes dried press residue, Hydrilla spp. dried press residue, and C. dactylon immature hay. All were experimentally harvested, and processed into pellets which were reground and blended into balanced pelleted diets containing the test materials as 33% of the organic matter. The E. crassipes plants were medium in size, flowering, with at least 80% float petioles and not washed before machine processing. Table 9 shows that the dried C. dactylon plant material had a higher organic matter content, as % of dry matter, because of its' lower inorganic matter content of 5% as compared to 28% in Hydrilla and 11% in E. crassipes. The crude protein content expressed as % of organic matter was above 10% in all test materials. Table 9. Composition of dried C. dactylon hay, Eo crassipes press residue and Hydrilla spp. press residue C. dactvlon E. crassiEes Hydrilla sEE-. Dry matter, fresh, % 5 11 Dry matter, as fed, % 92 90 92 Organic matter, 90 D.M. 95 89 72 Ash, % D.M. 5 11 28 Crude protein, % D.M. 10 10 13 Voluntary daily intake of dry matter and crude protein was apparently reduced by the higher ash content of the aquatic plant diets. See Table 10. The same finding has been reported in all subsequentresearch with aquatic plant residues. The daily intakes Table 10. Voluntary daily intake of diet components C. dactylon E. crassipes Hydrilla sEP Dry matter, g/kg BW* Organic matter, gjkg BW* Crude protein, g *Live bodyweight of test cattle 24 21 17 23 20 14 720 650. 520 of dry matter and crude protein were adequate for maintenance of bodyweight and positive nitrogen balance with all test diets. Table 11 shows that the digestion coefficient for dry organic matter was lowest in the diet containing E. crassipes. The digestion coefficients 47 411


for crude protein were lower in both diets containing aquatic plants. If digestion coefficients for crude protein were calculated by difference for the organic plant materials, it would be apparent that Table 11. Digestion coefficients for diet C. dactylon E. crassipes Hydrilla spp. 71 Organic matter, % Crude protein, % Cellulose, % 72 65 37 66 .52 31 48 54 they provided none of the digestible protein. This finding substantiated the belief that the press juice received most of the useful protein. Apparently, dry organic matter and cellulose is more digestible in Hydrilla than in E. crassipes residues and both are inferior in dry organic matter digestibility to C. dactylon immature hay. Further research on the nutritional value of dried E. products was not justified until a corrnnercially feasible processing method was perfected. Bagnall (1971) has published data on all aspects of the mechanical processing of these products. To answer the question "What is the market value of E. crassipes dried press residue as a component of cattle feeds?", a 112-day feeding trial was conducted with individually-fed steers to compare it with two popular competitive products, cottonseed hulls and sugarcane bagasse pellets, as the only source of bulky large particles in a high-concentrate cattle finishing Table 12 shows the ingredient composition of the test diets. Table 12. Ingredient composition of cattle finishing diets to compare the value of three sources of bulky, large particles Corn, yellow, steamed rolled Citrus pulp, press residue, dried Soybean oil meal, 50% C.P. Urea, 45% N Molasses, cane, standard Salt, trace mineralized Salt, common white Dicalcium phosphate Vitamin premix Test materials In ]SJlld!SJiih DI I E __ 48 % 50 20 7 1 10 0.5 0.5 1.0 10


., All 'criteria measured showed E. crassipes whole, chopped, wet pressed residue .to have a replacement value of at least equal to the competitive products, cottonseed hulls and sugarcane bagasse pellets. The epithelial lining of the forestomachs of cattle fed the diet E. crassipes was distinctly cleaner, and the rumen papillae exhibited less encrustation, clumping, hair matting and irregularity in structure. No-liver abscesses, enteritis or other abnormalities were found in the organs of the five cattle fed the containing E. crassipes and all of the carcasses u.s. Choice. This small study would indicate that the market value for low quality experimentally processed E. crassipes dried press residue might be based on its use as a replacement for cottonseed hulls and sugarcane bagasse pellets in cattle diets. Because commercial drying facilities were inadequate, research was centered on the processing of E crassipes wet pressed residue into silage. The fantastic water retention properties of the plant fibers allowed it to ,be transported without runoff of liquids at ,a dry matter content ranging from 12 to 15 percent., Bagnall. (1973) conducted extensive studies on all engineering as pects of the ensiling process. Research in progress (Baldwin, 1973; Byron, 1973) has shown that good quality silage can be produced from E. crassipes whole plant chopped wet pressed residue, fiX ash 16% of dry matter, when dried citrus pulp is added at a level of about 25% of the dry.matter. -This level. is exactly like that recommended years ago for preservation of freh chopped grass; namely, 200 pounds of dried citrus pulp per ton of fresh chopped forage. It is questioned whether the addition of sugarcane molasses at levels up to 1% of the ensiled material has been beneficial. It is apparent that excesses of inorganic matter in the freshly harvested material area barrier to rapid and sustained fermentation with production of desired organic acids. The pH of the ensiled material was not only correlated with ash content of the fresh plant material but also Kas correlated with acceptability of the silage by cattle. The first large-scale, commercial production of E. crassipes ensilage occurred in A commercial aquatic plant harvesting firm, Sarasota Weed ,. & Feed Co., Sarasota, Florida, was contracted to harvest E. crassipes from Paynes Prairie State Park, Gainesville, Florida. wet pressed residue was transported by truck to the University of Florida campus and moved by chain conveyors into a concrete tower silo. The success of the first venture should encourage investigation of the use of sealed silos located adjacent to large sources of water hyacinths. One obvious use is near waste disposal ponds in urban areas where E. crassipes crops can be grown to remove pollutants from the waste water then be mechanically harvested and ensiled in a seal silo which has a bottom unloader for the marketable silage. The cost of a permanent silo structure seems unjustified in remote areas; therefore, research was initiated on methods of preservingE. crassipes chopped, wet pressed residue in above-ground horizontal stacks. Preliminary studies (Byron, 1973) with various 49


organic acids in the preservatives have yielded silage from pilot silos which was clearly preferred by cattle over other treatments. Also spoilage was reduced by the acid preservatives. Further research on the ensiling of E. crassipes in stacks appears to be justified. 50


Classification of Water Hyacinth Products 1 2 According to International Feed Nomenclature Origin Water hyacinth Species Part Process Eichhornia crassipes Whole Whole w/preservative added Aerial part Leaves Float petiole Fresh Washed Chopped Ground Wet pressed Extruded Stolons Roots Process residue Process juice Process juice residue Elongated petiole Dehydrated Fan air dr ied Pellet.ed Ensiled, tower silo Ensiled, stack silo Stage of Maturity (Omitted, descriptive term substituted) % float petioles -::-;0--Size of plant (midget, small, mediumJ large, giant) Cutting Grade Example Cut 1 Max % ash Mn % C.P. Water hyacinth, Eichhornia crassipes, whole, washed, chopped, wet pressedreeidue, fan air-dried, groTInd, pelleted, reground, cut 1, 10% float petioles, max 25% ash, mn 12% C.P. lHarris, L.E., et aI, Utah Agr. Exp. Sta. Bul. 479, Nov. 1968. 2 Prepared by Dr. James F. Hentges, Jr., Professor of Animal Nutri-tion, Institute of Food and Agrc. Sciences, University of Florida, Gainesville, Florida, 32611. 51


.. ACKNOWLEDGEMENT The authors are indebted to the state of Florida, Department of Natural Resources, Bureau of Aquatic Plant Research and Control and to the Southwest Florida Water Management District for the financlal contributions and counselling in the prosecution of this project. These two state agencies funded that portion of the research that was performed by Drs. Shirley and. Hentges. The assistance of these agencies was not only greatly appreciated by the principal investigators, but also by the Director of the Florida Water Resources Research Center 52


REFERENCES CITED Army Corps of Engineers. 1946. Water hyacinth obstructions in the waters of the Gulf and South Atlantic States, 85th Congress, 1st Session, House document 37, U.S. Gover:nment Printing Office. Washington, D.C. Baldwin, John A. 1973. ruminant diets. ville, Florida. Utilization of ensiled water hyacinths in M.S.A. Thesis, University of Florida, Gaines-Bagnall, L.O., T.W. Casselman, J.W. Kesterson, J.F. Easley, and H. F. Hellwig. 1971. Aquatic forage processing in Florida. Paper 71-536, Proceeding Amer. Soc. Agri. Engr. Bagnall, L.O. 1973. Aquatic weed utilization: harvesting and processing. Froc: Weed Society, Atlanta, Georgia. Boyd, C.E. 1968a. Fresh-water plants: A potential source of protein. Econ. Bot. 22:359. Boyd, C.E. 1968b. Evaluation of some common aquatic weeds as possible feedstuffs. Hyacinth Control J. 7:26. Boyd, C.E. 1969. The nutritive value of three species of water weeds. Econ. Bot. 23:123. Boyd, C.K. 1970. Vascular aquatic plants for mineral nutrient removal from polluted waters. Econ. Bot. 24:95. Byron, H.T., Jr. 1973. Nutritive value of ensiled water hyacinth press residue for ruminants. M.S.A. Thesis, University of Florida, Gainesville, Florida. Chadwick, M.J. and M. Obeid. 1966. A comparative study of the growth of Eichhornia crassipes Solms. and Pistia straitiotes L. in water-culture. J. Ecol. 54:563. Combs, G.E., Jr. and H.D. Wallace. 1973. Use of dehydrated water hyacinth in swine diets. An. Res. Rpt. AL. 73-3, University of Florida, Gainesville, Florida. Datta, R.K. P.R. Chakrabarty, B.C. Guha and J.J. Ghosh. 1966. Studies on leaf 'proteins-preparation of protein concentrate from leaves of water hyacinth. Science and Culture. 32:247. Denton, J.B. 1967. Certain relationships between the chemical composition of aquatic plants and vlater quality .. Pro. 20th Ann. Meeting So. Weed Sci. Society. Harris, L.E. J.M. Asplund, and E.W. Crampton. 1968. An international feed nomenclature and methods for summarizing and using feed 53 r.;: I I r P" P n __ nrmiER .AG2F?'iiiiii;li!


data to calculate diets. Utah Agr. Exp. Sta. Bul. 479. Hentges, J.F., Jr. 1970. Processed aquatic plants for cattle nutrition. Proc. Aquatic Plant Res. Conf., University of Florida, Gainesville, Florida. Hentges, J.F., Jr., L.O. Bagnall, and R.L .. Shirley. 1972. Aquatic weed utilization nutritive value for livestock. Proc. Weed Sci. Society, Atlanta, Georgia. Hentges, J.F., Jr. and J.A. Baldwin. hyacinth products according to ture .. Unpublished, University Florida. 1973. Classification of water international feed nomenclaof Florida, Gainesville, Holm, L.G., L.W. Weldon, and R.D. Blackburn. 1969. Aquatic Weeds,. Science, 166:699-709. Lawrence, J.M. and W.M. Mixon. 1970. Comparative nutrient content of aquatic plants from different habitats. Proc. 23rd Ann. Meeting So. Weed SC.L. Society. Liang, J.K. and R.T. Lovell. 1970. Nutritional value of water hyacinth in catfish feeds. Hyacinth Control J. 20:44. Penfound, W.T. and T.T. Earle. hyacinth. Ecol. Monogr. 1948. The ,biology of the water 18:448. Salveson, R.E. 1971. Utilization of aquatic plants in steer diets: voluntary intake and digestibility. M.S.A. Thesis, University of Florida, Gairiesville, Florida. Shirley, 1970. Chemical content of aquatic plants. Proc. Aquatic Plant Res. Conf., University of Fiorida,Gainesville, Florida. Stephens, E.-L. 1972. Digestibility trials on ten elements and three toxicants in aquatic plant diets fed steers. M.S.A. Thesis, University of Florida, Gainesville, Florida. Steward, K.K. plants. 1970. Nutri.ent removal potential of various aquatic Hyacin'th Control J. 8 :34. Taylor, K.G. and R.C. Robbins. 1968. The amino acid composition of water hyacinth (Eichhornia crassipes) and its value as a protein supplement. Hyacinth Control J. 7:24. Taylor, K.G. 1969. The protein of water hyacinth (Eichhornia crassipes) and its potential contribution to human nutrition. M.S.A. Thesis, University of Florida, Gainesville, Florida. Vetter, R.L. 1972. Preliminary te'sts on the feeding value for cattle of fresh and processed water hyacinths. A.S. Leaflet R1691 Iowa State University, Ames, Iowa. 54 ...


Weldon, L.W., R.D. Blackburn, and D.S. Harrison. 1969. Common Aquatic Weeds. USDA Ag. Handbook 352, 42 pages. Yount, J.L. 1964. Aquatic nutrient reduction -potential and possible methods. Proc. 35th Ann. Meeting Fla. Antimosquito Ass'n. 55 I