CIVIL ENGINEERING STUDIES SANITARY ENGINEERING SERIES NO. 26 REMOVAL OF FREE-LIVING NEMATODES BY RAPID SAND FILTERS By RONALD LEIGH PETERSON Supported By DIVISION OF WATER SUPPLY AND POLLUTION CONTROL U. S. PUBLIC HEALTH SERVICE RESEARCH PROJECT WP 00047 and TRAINEESHIP GRANT SP-24-64 Property of COLLEGE OF ENGINEERING DOCUMENTS CENTER UNIVERSITY OF ILLIfcGSS 157 GRAINGER LIBS 1301 WEST SPRINGF2ELD AVENUE URBANA, ILLINOIS 61801 USA DEPARTMENT OF CIVIL ENGINEERING UNIVERSITY OF ILLINOIS URBANA, ILLINOIS MAY, 1965 REMOVAL OF FREE-LIVING NEMATODES BY RAPID SAND FILTERS by RONALD LEIGH PETERSON Supported by Division of Water Supply and Pollution Control U. S. Public Health Service Research Project WP-00047 and a U. S. Public Health Service Traineeship Department of Civil Engineering University of Illinois Urbana, Illinois May, 1965 REMOVAL OF FREE-LIVING NEMATODES BY RAPID SAND FILTERS by Ronald L. Peterson, M, S Department of Civil Engineering University of Illinois, 1965 This study was initiated to determine the fundamental factors which allow free-living nematodes to pass through rapid sand filters „ The effect of changes in influent conditions was studied to aid in finding methods of increasing the removal of nematodes by filtration so as to eliminate their presence in finished water supplies The percent removal of Diplogasteroides sp„, one of the predominate species found in waste treatment effluents, was correlated with the variables of flow, influent total worm concentration, influent motile worm concentra- tion, and worm size. Studies were run at 20°C using a 2-1/2 inch plexiglass filter column. The sand was examined to study the distribution of the worms after filtration and after backwashing. Results at flows of 2, 4, and 6 gpm/sq ft showed approximately 98 percent removal of nonmotile worms, but an average of only 25 percent removal of motile worms. The motility of the worms seemed to be the most important single factor in worm removal „ Smaller, less motile worms were removed better than larger, more motile ones. Percent worm removal with high concentrations of nonmotile worms was higher than with lower concentra- tions of motile worms. Examination of the sand showed worm retention throughout the entire depth but with the highest concentrations near the top of the sand. Ill ACKNOWLEDGEMENTS The author wishes to thank all those members of the Sanitary Engineering faculty and staff who contributed their talents to the various parts of this work He would especially like to thank: Dr. R S Engelbrecht for his continued patience and guidance throughout the experimentation and writing „ Dr„ Jo Ho Austin for his patience, valuable suggestions, and attentions to detail Mr, Gerald Steiner for his diligence and dependability The Division of Water Supply and Pollution Control, U S Public Health Service, for the funds which made this research possible IV TABLE OF CONTENTS Page ACKNOWLEDGEMENTS iii LIST OF TABLES vi LIST OF FIGURES vii 1. INTRODUCTION 1 1.1 The Problem 1 1.2 The Nematode 2 1.3 Scope of Study 3 2. APPARATUS 8 2.1 Reservoir 8 2.2 Air Pump 10 2.3 Filter Column 13 2.4 Influent Worm Supply 15 2.5 Sampling 16 2.6 Tubing 19 3. OPERATION 20 3.1 Backwashing 20 3.2 Worm Cultures 20 3.3 Worm Suspension Preparation 21 3.4 Flow Measurement 24 3.5 Sample Analysis 24 3.5.1 Filtration 24 3.5.2 Staining 25 3.5.3 Motile Count 27 3.6 Sand Sampling 29 Digitized by the Internet Archive in 2013 http://archive.org/details/removaloffreeliv26pete TABLE OF CONTENTS (Cont'd) Page 4. COMPUTATIONS 31 4.1 Water Computations 31 4.2 Data Processing 32 4.3 Sand Computations 33 5. RESULTS AND DISCUSSION 34 5.1 Total Worm Concentration 38 5.2 Motile Worm Concentration 40 5.3 Removal of Nonmotile Worms 41 5.4 Percent Motility and Percent Total Worm Removal 41 5.5 Motility Loss 46 5.6 Effect of Worm Size 50 5.7 Worm Distribution in Sand 52 5.8 Backwashing 58 6. CONCLUSIONS 60 7. SUGGESTIONS FOR FUTURE RESEARCH 61 REFERENCES 63 APPENDICES 65 1 VI LIST OF TABLES Table Page Id Filtration Variables Studied 4 2..1 Filter Gravel Characteristics 13 5 1 Percent Worms Removed by Backwashing 59 Vll LIST OF FIGURES Figure Page 1„1 Percentage of Motile Diplogasteroides sp. Nematodes at Various Temperatures 7 2,1 Filter Apparatus 9 2„2 Effectiveness of Air Flushing in Reducing Air Binding at Water Temperature of 35°C 11 2»3 Effectiveness of Air Flushing in Reducing Air Binding at Water Temperature of 18°C 12 2*4 Filter Column Detail 14 2o5 Electrolytic Pump 17 2„6 Flow Divider 18 3 1 Procedure for Last Nine Filtration Experiments 23 3 2 Sampling Equipment 26 3 3 Counting Dish 28 3,4 Sand Sampling Locations 3.0 5„1 Worm Concentration vs Time 9 Flow Rate - 2 gpm/sq ft 35 5„2 Worm Concentration vs„ Time, Flow Rate - 4 gpm/sq ft 36 5«3 Worm Concentration vs„ Time, Flow Rate - 6 gpm/sq ft 37 5»4 Effect of Variable Influent Worm Concentration Upon Effluent 'Worm Concentration 39 5 5 Motile VJorms Out vs Motile Worms In 42 5„6 Worms Out vs. Motile Worms In 42 5o7 Removal of Nonmotile Worms In Experiments with All Dead Worms 43 5„8 Percent Worm Removal vs„ Percent Motility 9 Flow Rate - 2 gpm/sq ft 45 5 9 Percent Worm Removal vs„ Percent Motility, Flow Rate - 4 gpm/sq ft 45 Vila LIST OF FIGURES (Cont'd) Figure • Page 5„10 Percent Worm Removal vs„ Percent Motility, Flow Rate - 6 gpm/sq ft 47 5 11 Percent Worm Removal vs Percent Motility, All Flow Rates 47 5012 Effect of Flow Upon Percent Removal of Motile and Nonmotile Nematodes 49 5013 Effect of Percent Motility Upon Removal of Nonmotile Worms 49 5,14 Nematode Size Distribution in Influent and Effluent 51 5 15 Change in Worm Concentration with Sand Depth 53 5016 Distribution of Nematodes in Sand 55 5«17 Percent of Total Worms Below Depth vs, Total Worms Removed 57 1, INTRODUCTION 1.1 The Problem Nematodes in public water supplies first came to the attention of Sanitary Engineers as early as 1918 when Cobb (1) discovered their extensive occurrence in slow sand filters. The conditions in slow sand filters were ideal for nematode breeding and they did so with such prolif- icness that their excretions were suspected of causing taste and odor problems in the filtered water, The environment in a rapid sand filter, with its periodic back- washing, is less suited to nematode breeding. The most recent indication of a problem came in 1955 when Chang et al, (2) discovered live, motile nematodes in concentrates of treated Ohio River tap water. Treatment of this water had included coagulation, sedimentation, filtration and chlori- nation, A survey of the treatment plant showed the worms to be present throughout the plant. They were apparently removed to a great extent by coagulation and sedimentation, but only poorly by filtration. Data on this latter observation, however, were inconclusive because most worm concentra- tions were too low to yield statistically significant results, A later survey by Chang et al, (3) showed the occurrence of nematodes in the effluent of rapid sand filters in 16 of 22 supplies, Kelly (4) showed that nematodes were highly resistant to chlorine, as much as 20 mg/1 being required to kill adult worms in 2,5 hours, A significant observation by Chang (5) was that ingested pathogenic bacteria and viruses were completely protected by the nematode even though the worms were chlori= nated to the extent that 90 percent were immobilized. This, combined with the apparent widespread occurrence of the worms in city water supplies, led to some concern over the possibility of nematodes being "carriers" of pathogens, protected from exposure to chlorine (in a water supply) and the subsequent possibility of infection of users of the water „ 1 2 The Nematode Many genera of nematodes are of concern in water treatment. In his survey of 22 water supplies, Chang (3) found the genera Monhystera , Aphelunchus , Rhabditis , Diplogaster , Chephalobus 8 Turbatrix, and D orylaimus ,, in order of decreasing occurrence, in finished waters , Engelbrecht et ale (6) observed Diplogaster , Diplogasteroides , and Rhabditis in order of decreasing occurrence in Illinois surface waters „ Engelbrecht et al. (6) found that waste treatment plants were "the origin of high nematode populations, while lower concentrations are contributed by surface and subsurface drainage," In their study, nematodes were found to persist in a stream for a considerable distance below the sewage treatment plant. The concentrations over a four-month period averaged over 1000 worms per gallon at a point 7,7 miles downstream of the sewage treatment plant. Nematode morphology has been completely covered by others (7, 8), but enough should be mentioned here to illustrate the problem. The worms are elongated, with a circular cross-section, a blunt head, and a tapered tail, Some species grow to a length of 4 mm with a diameter of 250 microns. The young and other species range up to this size. The generation time is measured in days (9), The adults and most larval stages of the free-living forms show a high degree of motility characterized by a rapid, undulating movement, This motility is adversely affected by extremes in temperature and pH (9) 1 3 Scope of Study All of the variables in rapid sand filtration could conceivably have some effect on the removal of nematodes „ However, some of these, such as turbidity and pretreatment , if included, would tend to obscure other effects of interest, such as those due to worm concentration and motility o For this study, therefore, only tap water was used for the filter influent o Nematodes obtained from cultures were injected into this flow and the removal characteristics of the filter studied „ The tap water was ground water which had been previously subjected to lime-soda soften- ing, filtration, and chlorination by the Champaign-Urbana water treatment plant o The variables investigated in this study and their ranges are shown in Table 1 1„ In all cases the ranges of the variables were selected to conform as closely as practicable with existing conditions in practice „ The choice of some, however, deserve comment „ Sand characteristics are probably an important variable and will be investigated in the future „ For these studies, however, a widely used sand with a uniformity coefficient (UC) of 1„5 and effective size (ES) of 0o5 mm was the only one used. The sand used was obtained from Muscatine, Iowa„ Periodically during experimentation the sand was removed and checked for grain size and distribution. If the uniformity coefficient (UC) was within ±00 05 units and the effective size (ES) was within ±00 02 mm of the desired values, the sand was carefully replaced and used again „ If not, new sand with the same specifications was prepared and used« A sand depth of 24 inches was used exclusively after preliminary tests at various depths disclosed that approximately 90 percent of the worms removed were removed by the top 2 cm of sand regardless of depth „ From this it was concluded that sand depth was not an important variable and a commonly used depth of 24 inches was chosen (10, 11 ) TABLE 1,1 FILTRATION VARIABLES STUDIED Variable Range of Study Flow Rate Sand Characteristics Sand Depth Filtration Time Influent Worms Concentration Total Motile Genera Life Stage Length Temperature 2, 4, and 6 gpm/sq ft (4 9, 9.8, and 14 „ 7 cu m/sq m/hr) UC = 1.5, ES = 0.5 mm 24 inches (51 cm) 8 hours 10-50/liter 0-25/liter Diplogasteroides Larvae and Adults 0.16 - 1.2 mm 20°C A filtration time of 8 hours was used for several reasons. Eight hours represents a normal minimum filtration time in practice; some plants prefer to backwash every shift for simplicity of operation (12). The experimental apparatus used required continual attendance during filtration, making experiments longer than 8 hours difficult. Eight hours was more than sufficient time to establish a trend , if there was one, or to attain an equilibrium with respect to effluent worm concentration „ The worm concentration in the influent water to the filter was sufficiently high to give a realistic and statistically meaningful result without being unduly difficult to count „ Chang et^ al (3) recommended that remedial measures be taken should the worm concentration in the finished water exceed 10/gal (about 2/liter) The influent concentrations used in this study generally yielded effluent concentrations of this mag- nitude or higher o No influent concentrations greater than 36 worms per liter (13) were found in the literature but it was felt that to properly evaluate the effect of concentration , values of at least 50 worms per liter should be usedo The motile count, of course, could vary anywhere up to the total count In actuality it was found difficult to get percentages of motility as high as 50 percent in the influent „ This was mostly due to the types of cultures used The number of generations required to produce a suffi- cient number of worms was high enough to result in an appreciable number of old, dead 9 worms being present „ Also, there was probably some loss of motility by the worms in passing through the apparatus before reaching the filter o Diplogasteroides sp „ was chosen for these studies because it is common to Illinois surface waters (6) and is classified in the same sub- family as Diplogaster , a genera commonly found in water supplies by Chang et al o ( 3 ) „ One other genera, Tricephalobus , was used Tricephalobus is smaller than Diplogasteroides and moves more slowly „ Comparison of the removal characteristics of these two genera permitted evaluation of the relative importance of motility and size Initial studies showed motility to be more important than size„ For this reason motility was taken as the independent variable , while the size distribution was used as found in the culture flask „ The lengths of worms used ranged up to approximately 1<>2 mm, with a predominance of the worms with length about „2 mm Temperature, of course, affects water viscosity and consequently head loss,, More important, however, is the effect of temperature on worm motility „ Chaudhuri (9) found the relationship shown in Figure 1 1<, Twenty degrees C was used in all cases because this yielded maximum motility with Diplogasteroides sp 100 _ CO w o f- < u 55 W ►4 M H O O w o < 55 W o « w &4 10 20 30 TEMPERATURE °C 40 FIGURE 1.1 PERCENTAGE OF MOTILE DIPLQGASTEROIDES SP. NEMATODES AT VARIOUS TEMPERATURES. (FROM CHAUDHURI (9)) 2 U APPARATUS The apparatus used in these experiments is shown in Figure 2„lo The over-all flow pattern is from the reservoir through the pump, "orifice meter," past the "division" manometer , and onto the filter This arrangement was determined as best only after considerable preliminary testing of its individual components , which are perhaps best explained separately „ 2 1 Reservoir The reservoir served 3 functions; 1) to maintain constant head 2) to maintain constant temperature 3) to equilibrate dissolved gases Since a centrifugal pump was used, a constant head was needed for constant flow. Constant head was maintained by an overflow from the reservoir. Hot and cold tap water was mixed in the reservoir for proper temperature and constant temperature was maintained by observing a ther— mometer and making adjustments to the hot and cold water flows as needed „ The size of the reservoir caused temperature fluctuations to be gradual and allowed time for correction „ When the temperature of water increases, the solubility of gases in the water decreases „ If the water contains much gas, a sufficient increase in temperature will cause some gas to come out of solution in the form of many small bubbles „ These bubbles, when lodged in the sand, were observed to cause air binding in the filter. D < w M CM D O M 10 In maintaining a temperature of 20°C S cold tap water at about 16° C was mixed with hot tap water , The warming of the cold tap water caused a considerable amount of dissolved gas to be released. Much of this gas escaped from the water during retention in the reservoir „ However, at 4 and 6 gpm/sq ft, insufficient detention time was provided and addi- tional measures became necessary to prevent air binding. These are explained in the next sect ion „ 2,2 Air Pump Air pumped through a sparger submerged in the water above the sand in the filter proved effective in preventing air binding. The flush- ing action of the large bubbles rising against the flow apparently "scrubbed" out the small bubbles being carried in it. The effectiveness of this technique is shown in Figures 2,2 and 2,3, Four tests were made, two at a water temperature of 35°C and two at 18°C At each temperature, one test was made with flushing, and one with- out. The air temperature was 27°C, At both water temperatures flushing eliminated air binding, as is shown by the head loss curves. With warm water without flushing, air binding occurred to such an extent that the flow could not be maintained after 35 minutes. The air pressure provided by the flushing served another function, At flows of 4 and 6 gpm/sq ft the head loss would require a longer filter column so that the water level inside could be raised. To avoid a longer filter column air pressure was used in lieu of the increased water head. The air head was regulated with needle valve at E (Figure 2,1), The total head above the filter was read with the manometer shown at H, Head loss 11 oes/ioj «31VH W01J seipux «SS0T (3V3H 12 o o 00 o & ac OS o 25 M as CO O 25 M o Q 25 M O 25 M ac CO U. OS O W w w 2: M E- o w CN 1 M fiu, 09S/TU1 •aiVH M01J saqouT *SS0T QV3H 1 I 1 13 was obtained by subtracting manometer reading I from manometer reading H. It should be emphasized here that flow was held constant through- out an experiment with head loss allowed to vary. Since only tap water and worms were being filtered, and air binding was minimized, head loss rise during an experiment was gradual and of a low magnitude,, 2 3 Filter Column The filter column used is shown in detail in Figure 2,4, The column was segmented so that it could be taken apart and the sand examined, Robeck et al. (14) mentions that wall effects in a filter column will be negligible as long as the column diameter is at least 50 times the media particle diameter. A diameter of 2 1/2 inches fulfills this require- ment o The ASCE manual on water treatment plant design (10) recommends a gravel bottom of the dimensions shown in Table 2,1, TABLE 2,1 FILTER GRAVEL CHARACTERISTICS Layer Depth Gravel Size (inches) (inches) 1 * 1-3/4 2 3 3/4-1/2 3 3 1/2-1/4 4 4 1/4-1/8 5 4 1/8-1/16 M at least 4 inches above wash water inlet Layers 1 and 5 were not used in the experimental filter. The larger gravel was felt to be unnecessary because the filter was sufficiently INFLUENT 2 1/2" I. D. PLEXIGLASS O-RING ■*8 X — T 3 O to 3 IT) 1 II II ELEVATION -~- EFFLUENT SECTION 14 FIGURE 2.4 FILTER COLUMN DETAIL 15 small that backwash water distribution was not critical <, The smaller sizes of gravel showed a tendency to mix with the sand during backwash and so were not used. Even with this size eliminated , the sand showed no observable tendency to escape through the gravel bed. The filter column therefore had 10 inches of gravel at the bottom, the sand starting at the lower flange. 2,4 Influent Worm Supply Possible variables in an influent worm supply ares 1) worm genera 2) worm size 3) worm size distribution 4) total worm concentration 5) motile worm concentration During any particular filtration period it would be desirable to hold all these factors constant. The first three presented little problem; only one genera was used and during a filtration period of only 8 hours, size and size distribution variation due to worm growth would not have time to occur „ Total worm and motile worm concentrations were more difficult to control. The method used was to pump a thoroughly mixed worm suspension through a capillary tube into the main flow at point A (Figure 2,1), To avoid a temperature shock which might effect the worm motility, the suspen- sion was kept at 20°C by a water bath using water from the reservoir. The pumping of the worm suspension was done with an electrolytic pump of a type described by Symons (15), The arrangement is shown in more 16 detail in Figure 2.5. The rate of flow from the worm suspension was dependent upon the production of hydrogen and oxygen from the electrolysis of water in bottle number 1, which in turn was regulated by the current from the D C, source. Thus, the rate of flow from the worm suspension was regulated quite easily by the rheostat. The pressure in the system was monitored with the manometer from bottle number 2. Tests with this apparatus showed that it would deliver a nearly constant total worm concentration and that the motility loss over an 8-hour period was negligible. 2.5 Sampling It was not possible to maintain the influent worm concentration exactly constant; so some means of measuring the variations was needed. Grab samples were not feasible because they required periodic interruption of the flow, upsetting the equilibrium in the filter. The method used was a continuous process, whereby a constant portion of the flow was diverted for a sample. The resultant composite was processed at about half-hour intervals and its concentration taken as the average worm concentration over the collection period of the sample. The critical part of the influent sampling procedure was the flow division. Division of the flow had to be such that the influent sample was representative of the influent to the filter. The means em- ployed is detailed in Figure 2,6, Since the branches of the inverted "Y" were practically identical it follows that, if the flows through them were equal, with complete mixing the worm concentrations would also be equal. Every attempt was made to keep the flows in the two branches equal. The flow division was monitored with the "division manometer." Keeping the G 3 3 1 17 • W o o PS • 3 Q o en p< u 1-1 e- >- o PS o w w in CM g CD M 3/16" I. D. PLEXIGLASS ♦ // /7 nA /.// 60 c w w 18 3/8" I. D. PLEXIGLASS // // // v // \ 3/16" I. D. PLEXIGLASS /y // ov ■3/8" I. D. RUBBER VALVE F "DIVISION MANOMETER" COLORED CC1, FIGURE 2.6 FLOW DIVIDER I 19 head difference constant between the two branches assured constant flow division. Any adjustments required were made with the valve at F (Fig- ure 2,1). All of the filter effluent was diverted to the effluent composite sample, yielding an effluent sample volume equal to that of the influent sample. The validity of this sampling procedure was tested with an experiment without sand in the filter „ Obviously, with this arrangement the average worms per liter in the influent should equal the average worms per liter in the effluent. The values obtained from a 4 1/2-hour period were 122.2 average worms per liter in the influent against 122.8 average worms per liter in the effluent, 2.6 Tubing The tubing used was tygon or rubber, depending upon suitability. The size ranged from 1/2 inch I.D. at A to 1/4 inch I.D. above the filter. A progressive decrease in tubing size was found to alleviate air bubble problems by progressively accelerating the flow. 20 3. OPERATION During operation of the apparatus, every attempt was made to minimize variations in flow, influent worm concentration, and temperature , A typical experiment required about 10 hours and consisted of the following steps s 1) backwash filter 2) prepare worms 3) attain flow equilibrium M-) start worm injection 5) take samples approximately every half-hour for 8 hours If the sand was to be examined, this was done immediately upon drainage of the filter after termination of the experiment „ 3 1 Backwashing Since the same sand was to be used repeatedly, it was necessary to clean it thoroughly before every experiment. Examination of the sand after backwashing for 30 minutes at 50 percent expansion disclosed no worms. Therefore, this procedure was adopted for sand cleansing, 3„2 Worm Cultures The worms were grown in the same medium used by Chaudhuri (9), Three ml of this medium in a T-30 culture flask generally yielded a total population of about 10,000 worms (adults and larvae) after three weeks of incubation at 20°C, Since further growth after three weeks resulted in overcrowding and an undesirably high percentage of dead worms , only cultures up to 3 weeks old were used for these studies. 21 3,3 Worm Suspension Preparation The product of flow rate times worm concentration times filtration time yields the total worms incident upon the filter „ Since the flow is divided in half above the filter, twice this number yields the total worms required in the worm suspension bottle for a given experiment,, As mentioned in section 3.2, the number of worms in a culture flask after 3 weeks' incu= bat ion at 20°C was usually about 10 s 000. Knowledge of this number and the calculations described below then yielded the number of flasks needed for the required number of worms . To calculate the number of flasks required for, say, 2 gpm/sq ft and a worm concentration of 50 per liter; 1) flow rate = 4.3 ml/sec = 15.48 1/hr 2) total worms required = 15.48 1/hr x 50 worms/1 x 8 hours x 2 = 12400 worms 3) total flasks required = , _ .,.. 1 T-30 flask at 3 weeks 1 _ c „ , 12,400 worms x ■ ■ ■ — y~ ■ .. = 1.25 flasks . ' 10,000 worms To a close approximation; use 1 flask for 2 gpm/sq ft use 2 flasks for 4 gpm/sq ft use 3 flasks for 6 gpm/sq ft The worm feed suspension was prepared while the filter was being backwashed. To obtain variable fractions of motile worms with minimal variations in total concentration, worm size, or size distribution, the following procedure was used, The required number of worms for one experi= ment was calculated as described above . This was then multiplied by three and the required number of culture flasks estimated. These flasks were then washed into a 50-ml beaker with about 35 ml of water. The beaker was stirred on a magnetic mixer and 3-10 ml portions withdrawn and placed in 22 separate test tubes „ One of these was washed into the worm concentrate bottle and the bottle filled with distilled water and placed on the apparatus „ The second was placed in a refrigerator,, A sufficient amount of a 10 percent solution of Eosin-Y dye was added to the third tube to yield a 2 percent solution,, The first tube, used immediately, yielded the highest percentage of motile worms; the second, used after two or three days in the refrigerator, yielded a somewhat lower percentage; the third tube 9 used after four to five days contact with the dye s yielded only dead worms „ For the last nine experiments a slightly different procedure was used. The objective of this series was to obtain data on worm removal at flow rates of 2, 4, and 6 gpm/sq ft and various motile worm concentra- tions with all other factors held as constant as possible „ To hold total worm concentration constant, three times as many worms were used at 6 gpm/sq ft as were used at 2 gpm/sq ft„ Enough worms were collected to allow this procedure (see calculations below) and the experiments were made on successive days,, The worms were kept in a refrigerator during the entire period and the normal attrition of the motile worms was sufficient to yield gradual variation of the motile worm concentration „ The experi- mental procedure is shown more clearly by a flow diagram (Figure 3„1) The previously used procedure was similar to this except that only three experiments were conducted in sequence with any particular worm culture and all were at the same flow rate with all dead worms used for the last One advantage of both these procedures is that they afford a cross-check on the precision of the sampling procedure „ With the methods used the total worm concentrations in any particular series whould be approximately equal (see Appendix A)„ I 23 18 T-30 FLASKS AT 3 WEEKS WORM STORAGE IN REFRIGERATOR DAY 1 DAY 2 DAY 3 DAY 4 DAY 5 DAY 6 DAY 7 DAY 8 DAY 9 10 ml 30 ml 20 ml 10 ml 30 ml 20 ml 10 ml 30 ml 20 ml EXPERIMENT AT 2 GPM/SQ FT EXPERIMENT AT 6 GPM/SQ FT EXPERIMENT AT 4 GPM/SQ FT EXPERIMENT AT 2 GPM/SQ FT EXPERIMENT AT 6 GPM/SQ FT EXPERIMENT AT 4 GPM/SQ FT EXPERIMENT AT 2 GPM/SQ FT EXPERIMENT AT 6 GPM/SQ FT EXPERIMENT AT 4 GPM/SQ FT STAINING AND COUNTING FIGURE 3.1 PROCEDURE FOR LAST NINE FILTRATION EXPERIMENTS 24 ments: With 3 experiments at each flow and same culture for 9 experi- 1) total flasks required =3x(l+2+3)=18 flasks 2) dilute to 190 mis 10 ml from beaker for 2 gpm/sq ft 20 ml from beaker for 4 gpm/sq ft 30 ml from beaker for 6 gpm/sq ft 3«4 Flow Measurement During filtration,, the flow was continually monitored with an orifice meter and differential manometer „ Twice an hour, in conjunction with the collection of the samples 9 both the influent (sample flow) and the flow from the filter were measured „ Every attempt was made to keep both flows constant „ The maximum percent mean deviation in the flow values measured was 5o45 percent , but the average was only 1„39 percent „ The flow was measured by collecting it in a 2-liter graduated cylinder and timing with a stop-watch over about 2 2-minute period „ This graduate was then emptied into a composite container and the cycle repeated with the next sample. Any required adjustments were made with the valve located at either B or G (Figure 2 1), depending upon whether the total flow or flow division was in error 3 5 Sample Analysis 3o5 l Filtration The worms were separated from the water by filtration through 5-micron membrane filters, a procedure established by Chaudhuri (16) At the higher flow rates the volume of composite sample obtained made it impractical to filter the entire quantity; so a method for 25 obtaining a reliable aliquot was devised Figure 3 2 shows the equipment usedo A 5=gallon carboy containing the composite was stirred vigorously , using a magnetic mixer The vacuum of the filter apparatus was used to draw the water up a tube and through the filter „ The volume of the sample filtered was measured by graduations on the vacuum flask The precision of this technique was tested on three samples by withdrawing four 2-liter aliquots from each Statistical analysis using the t-test gave an average 95 percent confidence interval equal to *12 D 7 percent of the mean In another test three 4-liter aliquots yielded a 95 percent confidence interval equal to *6„7 percent of the mean This showed that larger sample size was preferable and that when using a 4-liter sample a value for worm concentration within about *7 percent of the mean could be expected in 95 percent of the samples „ Effect of staining was not a factor In these tests because only all dead, completely stained worms were used; no staining was necessary between sampling and count Ing At all flows 9 therefore, 4 liters of the composite were filtered to obtain a sample for determination of worm concentration „ Actually, the total volume withdrawn from the composites was 8 liters, since 4 liters were filtered for a motile count and another 4 liters for a total count At 2 gpm/sq ft the volume of composite obtained was just sufficient to allow this procedure „ 3o5 2 Staining Chaudhuri et al. (17) showed that 100 percent staining of nematodes was achieved after 72 hours' contact with a 2 percent solution of Eosin-Y dye This staining of the nematodes greatly facilitates countings 26 3/8" I. D. TYGON TUBING 5-GALLON CARBOY 3/8" 0. D. PYREX TUBING IN RUBBER STOPPER 5 u MEMBRANE FILTER TO VACUUM PUMP 4-LITER VACUUM FLASK FIGURE 3.2 SAMPLING EQUIPMENT 27 especially in samples with high numbers of worms „ Therefore, to obtain a total count s a 4— liter aliquot from the composite sample was filtered and the membrane placed in a test tube and covered with a 2 percent Eosin-Y dye solution,, After at least 3 days the contents of the test tube were washed and filtered again „ This membrane was then placed with the retained material against the grid marks of a plexiglass counting dish (Figure 3,3), A drop of water on the counting dish was found sufficient to hold the membrane in place , The dish was then inverted and the membrane examined through the bottom of the dish with a dissecting microscope at a magnifi- cation of 24X 3,5,3 Motile Count To obtain a motile count, 4-liter aliquot s from the composite were filtered and the membrane washed into a counting dish. The contents of the dish were examined immediately under a microscope and the motile worms counted. For this investigation a motile worm is defined as one which could be seen to be obviously moving under cursory inspection „ Motility was believed to be of importance because it might allow the worms to wiggle through the sand bed where they would otherwise be retained „ Accordingly,, it was felt that if the motility was too slight to be obvious upon exami- nation, it would be too slight to be effective in the filter and therefore should not be counted,, While this is admittedly an approximation, it is apparently the only one feasible. It is important to note here that the motile count obtained is not the same as a viable count, which would probably be much higher. 3 3/4" 28 PLAN MICROSCOPE FRONT VIEW CO MICROSCOPE FILTER PAPER t POSITION FOR TOTAL COUNT POSITION FOR MOTILE COUNT FIGURE 3.3 COUNTING DISH 29 3 g 6 Sand Sampling To evaluate the removal mechanism , it was necessary to examine the sand after filtration „ After preliminary examinations of the sand showed that most of the worms would be found in the upper layers s the samples were taken as shown in Figure 3 e M-« Immediately after filtration, the sand was drained and samples taken „ Each sample consisted of a care- fully scraped layer 0,1 cm deep. This sand was weighed as quickly as possible to avoid evaporation errors and an accurately weighed 2 to 3 gram aliquot was removed,, To isolate the worms from the sand, this aliquot was treated according to the method devised by Baliga (18), using centrifugal flotation in a sugar solution. The worms were then stained and counted by the usual procedure. Since the weight of the sand sample was known, the worm concentrations could be expressed as number of worms per gram of wet sand 30 NUMBER OF WORMS/GM OF SAND SAMPLE DEPTH (cm) THICKNESS (cm) 1 .1 0.1 0.1 - 2 3 4 0.1 4. 12 0.1 5 28 0.1 O O Q < p. 60 0.1 FIGURE 3.4 SAND SAMPLING LOCATIONS 31 4. COMPUTATIONS 4.1 Water Computations The data obtained from an experiment consisted of the following series of observations against time; 1) flow 2) head loss 3) influent, total worms per liter 4) influent, motile worms per liter 5) effluent^ total worms per liter 6) effluent, motile worms per liter. These data were taken at different times and at various intervals, approxi- mating a half hour. From these data the following values were computed? 1 2 3 4 5 6 7 8 9 10 11 12 13 average flow average head loss average worms per liter in average motile worms per liter in average worms per liter out average motile worms per liter out total worms in total motile worms in total nonmotile worms in total worms out total motile worms out total nonmotile worms out total worms removed 32 14) total nonmotile worms removed 15) total motile worms removed 16) percent worms removed 17) percent motile worms removed 18) percent nonmotile worms removed 19) percent motile worms in influent 20) percent motile worms in effluent , The computational procedure was as follows „ For flow and head loss, two adjacent values were averaged and multiplied by the time period between them,, This was done for all the values and the sum of the products divided by the total time,, The procedure was the same for the worm data except that adjacent concentrations were not averaged; each concentration was assumed to be the average over its sampling period „ By the above procedures the data obtained was time-weighted, thus eliminating the necessity for obtaining data at exactly equal intervals „ The computations involved were not profound; but to relieve the tedium, and to insure accuracy, the data were processed using an IBM 7094 computer „ The computer was also used to calculate the range of each set of values, the mean deviation, and percent mean deviation , Summaries of the data obtained from each experiment are shown in Appendix A„ 4 2 Data Processing Once the data for each experiment had been obtained, it was necessary to evaluate the interrelationships between them There was a computer program available to compute means, standard deviations, correla= tion coefficients and a least squares straight line fit for any assemblage 33 of data The 20 values previously computed for each test were processed by the computer according to each flow rate and in aggregate (i e„, over all flows )<, The correlation coefficients between the various data were then examined to discover the significant relationships „ Various transformations of the data (arithmetic-log, log-log , reciprocal-arithmetic) to determine the best fit to other than a straight line were also tried,, If a correlation coefficient significantly higher than that for a straight line was obtained, the data were plotted and pre- sented as indicated. The arithmetic correlation coefficients are shown in Appendix B 4 3 Sand Computations Data obtained from examination of the sand consisted of s 1) depth of sample 2) weight of sample 3) thickness of sample 4-) weight of sample aliquot 5) number of worms in aliquot „ From these values the number of worms per gram at each depth could be computed o Knowing the distance between samples, then, allowed calculation of the total worms in the sand. 34 5 RESULTS AND DISCUSSIONS In Figures 5„1, 5„2, and 5„3 results of typical experiments at flows of 2, 4, and 6 gpm/sq ft are plotted, respectively „ From these plots it is apparent that time has little effect upon the worm removal characteristics of the filter; the effluent worm concentration after the first hour is essentially the same as after 8 hours; any changes are attributable to variations in influent worm concentration Study of these variations, however, is unrewarding, except in a general sense, because they are of a low magnitude and too sporadic „ Generally, peaks in the influent worm concentration are reflected by more subdued peaks in the effluent worm concentration „ And, generally, a greater proportion of the motile worms than of the nonmotile worms pass through the filter For more consistent and specific results it is preferable to speak in terms of averages over the entire filtration period of about 8 hours (as in average worms per liter in), or in terms of totals (as in total worms in)„ One is calculable from the other; the form best used is dependent upon which most clearly presents the data This procedure is valid and desirable because of the character of the variations during the filtration period,, Consistency of worm removal during an experiment was not unexpected o Only tap water was used; there was little in the water to change the characteristics of the sand„ This is reflected in the fact that head loss increase over a filtration period was generally less than 10 percent o Considering this, and provided that the worm concentration in the influent was not so high as to produce overloading of the filter sand, i iQ i r-^i 1 \ \ , \ V * / 1 / 1 *-* 1 S5 1 1 •J 1 1 55 \ M 1 1 1 1 \ \ \ \ \ \ \ \ - > t - / i / / / / <* « ■ o \ DC i - CO » J 32 I < <* i o o g 4 •J I— I ' r-t . O I £ 1 W 1 <) 1 1 1 1 / i / / * - / / / 1 1 i 1 1 I o o to o m o a- o CO o CN O O CM 3 « £ £ CO CN Cm ni aain H3d swhom d- CO o s w CJ o o OS o 5= CD to w :s o w as > o W O O a OS o 3: CO a H NI H31I1 H3d SWHOM xno aain Had swhom 38 there is no reason to expect substantial variance in the percentage of worms removed during an 8-hour filtration period „ 5»1 Total Worm Concentration In spite of the previous statements about the desirability of con- siderations of averages and totals over an entire experiment, there was one experiment where study of variations within the filtration period was re- warding o In this experiment (Figure 5„4) the tube leading from the worm suspension bottle became clogged after about 2 hours* filtration and produced extreme variations in influent worm concentration,, In addition, HC1 had been inadvertently used instead of H SO in the electrolytic pump The result was the generation of chlorine gas which, apparently, gradually killed the motile worms in the worm suspension bottle, resulting in a gradually decreasing in- fluent motile worm concentration „ When the run was terminated after 5 hours, the pH in the suspension bottle was found to be 3 lo This experiment was useful, however, precisely because of these deviations „ Comparison of the influent and effluent total worm concentrations shows that the wide variations in the influent had much less effect upon the effluent o A rise from 16 worms per liter to 32 worms per liter in the in- fluent produced only a slight rise in the effluent total worm concentration 9 as shown by the moving average „ This average was computed using successive groups of 3 adjacent values; the middle value in each group had twice the weight of those on either side,, The effect of a change in the influent motile worm concentration on the effluent, however, was quite marked The effluent motile worm con- centration closely paralleled the influent motile worm concentration The same sort of effect is shown in Figures 5„1, 5 2, and 5„3, although it is 39 40 30- w M ►J PS W20 Oh w O 10- — 1 1 1 i i _ - INFLUENT - - ^-^q/^^^- TOTAL WORMS - A. "■ / n .MOTILE WORMS A Y^ A / — ""A. 1 1 1 =^=A— -L-A A L H 20 O PS w M •J to o TOTAL WORMS A— A-- t r EFFLUENT ---&--.. A- i -A 2 3 TIME, hours FIGURE 5.4 EFFECT OF VARIABLE INFLUENT WORM CONCENTRATION UPON EFFLUENT WORM CONCENTRATION 40 not as apparent because the variations are of a much lesser magnitude „ 5 2 Motile Worm Concentration From the above observations one would conclude that motile worm concentration was a more important variable than total worm concentration From correlation studies between averages over 8 hours for the various experiments at various flows, the correlation coefficient (r) between average motile worms per liter out and average motile worms per liter in was found to be 0„994 at 2 gpm/sq ft, 998 at 4 gpm/sq ft and o 987 at 6 gpm/sq fto Over all flows the correlation coefficient was found to be 0o990 o This means that 98 percent" of the variations in one are directly attributable to variations in the other, regardless of either total worm concen tration , m otile worm concentration 9 or flow rate , over the ranges studied o The relationship between average motile worms per liter in and average motile worms per liter out for all flows is shown in Figure 5 5 According to the least-squares straight line of best fit for this graph, the average motile worm concentration in the effluent was about 87 „ 5 percent of that in the influent regardless of flow rate or motile worm concentration, This shows that most of the motile worms pass through the filter „ The chance of correlation coefficients as high as the above occurring accident" ally, for the amount of data collected and available for analysis, is less than one in a thousand (19) In Figure 5„5, as in all figures to follow, data at 2 gpm/sq ft will be designated by a small circle, data at 4 gpm/sq ft by a square, and data at 6 gpm/sq ft by a triangle „ In addition, those points representing data obtained from the last series of 9 experiments described in part 3 D 3 2 0o98 = o 99 41 are designated by solid shapes, the other data by open shapes „ The question now arises s "What effect does influent motile worm concentration have upon total worm concentration in the effluent?" These data are plotted in Figure 5 6 o Here the coefficient of correlation was 0„945 s again regardless of flow or total or motile worm concentration „ This means that 89 percent of the variations in effluent worm concentration were due to variations in the influent motile worm concent rat ion „ 503 Removal of Nonmotile Worms From the above discussion, the natural conclusion is that influent nonmotile worm concentration has little effect upon effluent worm concentration o Experiments conducted with all dead worms (see Appendix A, experiments 67, 74, 80, 83, 84, 89) showed percentages of removal ranging from 96 o 6 percent to 99 „ 5 percent with an average of 98 percent „ These data are plotted in Figure 5 7 It can be concluded from these data that: the percent removal of nonmotile worms was about 98 percent regardless of flow rate or worm concentration,, This substantiates the finding that effluent worm concentration was primarily dependent upon influent motile worm concentration, since an average of only 2 percent of the nonmotile worms ever appeared in the effluent „ 5 4 Percent Motility and Percent Total Worm Removal The average percent removal of nematodes during an 8-hour filtration period was found to correlate well with the percentage of mo- tile worms in the influent and with the influent motile worm concentration „ Correlation with the former (percent motility in the influent) was slightly 42 o w to o M O X. 8 ^u 1 l i i r = .99 A y 15 r 2 = .98 Sm - 10 i 'M x6 — SLOPE = 0.875 - 5, - Of i i i 5 10 15 AVERAGE MOTILE WORMS PER LITER IN FIGURE 5.5 MOTILE WORMS OUT vs. MOTILE WORMS IN o t-H ►J PS M Ph CO OS o w u a SLOPE =1.42 r = .945 r 2 = .89 5 10 15 20 AVERAGE MOTILE WORMS PER LITER IN FIGURE 5.6 WORMS OUT vs. MOTILE WORMS IN 43 12,000- 2,000 4,000 6,000 8,000 10,000 12,000 14,000 TOTAL NONMOTILE WORMS IN FIGURE 5.7 REMOVAL OF NONMOTILE WORMS IN EXPERIMENTS WITH ALL DEAD WORMS <+4 better, however This probably was because this factor takes into account the total worm concentration, which does have some effect since , from section 5„3, about 2 percent of the nonmotile worms appear, on the average, in the effluent „ In a plot of percent of total worms removed against percent of motile worms in the influent, at zero percent motility the percent of worms removed is that of nonmotile worms , At 100 percent motility, the percent of worms removed is that of the motile worms „ In Figures 5„8, 5„9, and 5 o 10 the percent of total worms removed during an 8-hour filtration period is plotted against the average percent of motile worms in the influent during that period for flows of, respec= tively, 2, 4, and 6 gpm/sq ft„ As stated in section loS^ motility percent - ages greater than about 50 percent were not obtained with the methods used A straight line fitted by least squares to these data (shown by the solid lines), when extrapolated, indicated zero percent removal of worms at motilities ranging from about 66 to 70 percent, depending upon the flow rate This situation was resolved by reasoning that the percentage removal of motile worms, as calculated from each experiment with motile worms, could logically be plotted at 100 percent motility, thus giving at least an idea of the location of the curve at greater than 50 percent motile worms in the influent „ Taking the average percentage of motile worms removed at each flow, the locations of the points suggested a curve of the types 1/PWR = b + a(PMWIN) where PWR is defined as the percent total worms removed, PMWIN is defined 45 100© 10 i 1 1 1 r 100/PWR = 1.005 + *030 PMWIN AVERAGE PERCENT- MOTILE WORMS REMOVED i _L I _L _L _1_ .L 10 20 30 10 50 60 70 80 90 100 PERCENT MOTILE WORMS IN INFLUENT (PMWIN) FIGURE 5.8 PERCENT WORM REMOVAL vs. PERCENT MOTILITY, FLOW RATE - 2 GPM/SQ FT 1001 10 1 1 1 T = .97 + .032 PMWIN AVERAGE PERCENT^" MOTILE WORMS REMOVED _L J_ JL _l_ _L _L _L J_ 10 20 30 40 50 60 70 80 90 100 PERCENT MOTILE WORMS IN INFLUENT (PMWIN) FIGURE 5.9 PERCENT WORM REMOVAL vs. PERCENT MOTILITY, FLOW RATE - 4 GPM/SQ FT 46 as the percent motile worms in the influent, and a and b are constants , The computer was used to find the curves of best fit at each flow and the results plotted as the dashed curves in Figures 5,8 S 5 9, and 5 o 10„ The differences between these curves are insignificant; it can therefore be concluded that the percentage removal of nematodes varied with the per- centage of motile worms in the influent 9 as shown by the dashed curve in Figure 5„11, independently of flow or total worm concentration This curve is the least-squares line of best fit to all the data at all flows and to the average percentage of motile worm removal over all flows. The conclusion that percent worm removal was independent of flow rate was unexpected. It seemed logical to suppose that the percent removal of worms at any particular percent motility would decrease with an increased flow rate, A possible reason for this behavior can be explained with reference to Figure 5,12, which shows flow plotted against the percents removal of motile and nonmotile nematodes for all experiments with motile nematodes present. It can be seen that the percent removal of nonmotile worms decreased slightly with an increased flow, but that the percent removal of motile worms increased. These two counter-acting factors apparently combined to yield' an over-all percent removal independent of flow rate, 5,5 Motility Loss The nonmotile worms behaved similarly to any other inanimate particle with respect to their decreased removal at increased flows 9 i e , the higher the flow rate= the greater the probability of one being carried through the filter. The unexpected behavior of the motile worms may be 100 10 i 1 1 1 1 i 1 1 r \ r 100/PWR = 1.034 + .030 PMWIN AVERAGE PERCENT MOTILE WORMS REMOVED J_ -L _L _L J_ _L 10 20 30 40 50 60 70 80 90 100 PERCENT MOTILE WORMS IN INFLUENT (PMWIN) FIGURE 5.10 PERCENT WORM REMOVAL vs. PERCENT MOTILITY, FLOW RATE - 6 GPM/SQ FT 47 100 10 ? i i i i i i i r = 1.01 + .030 PMWIN AVERAGE PERCENT MOTILE WORMS REMOVED _L J_ X JL JL _L _L 10 20 30 40 50 60 70 80 90 PERCENT MOTILE WORMS IN INFLUENT (PMWIN) 100 FIGURE 5.11 PERCENT WORM REMOVAL vs. PERCENT MOTILITY, ALL FLOW RATES 48 explained by considering that the increased turbulence and subsequent battering at higher flow rates caused an increased loss of motility by the nematodes „ A motile worm which lost its motility became a nonmotile worm, more easily removed by the filter „ Provided that the magnitude of the motility loss due to increased flow is greater than the decrease in removal of nonmotile worms due to increased flow, an increase in flow will result in an increased removal of motile worms . It was stated previously that, in the filtration experiments using all dead worms 9 the percentage removal of nonmotile worms was about 98 per- cent. Filtration experiments involving motile worms, however 9 invariably produced percentages of nonmotile worm removal less than 98 percent This is shown in Figure 5„13o As the percentage of motile worms in the influent increased, the percentage removal of nonmotile worms decreased, with little differences between the various flows , The explanation for this behavior is that the percentage of non= motile worm removal computed for these experiments is an apparent percentage removal computed by taking the difference between the total nonmotile worms in and the total nonmotile worms out It appears that what actually occurred was a loss of motility by some worms passing through the filter „ These worms, that had been counted as motile worms in the influent, were thus being counted in the effluent as nonmotile worms that had apparently passed through the filter, thus reducing the apparent percent removal of nonmotile worms . The exact magnitude of this loss of motility proved difficult to evaluate o Theoretically, if the percentage removal of nonmotile worms remained constant at 98 percent, then 2 percent of the nonmotile worms into Q 100 W > o 90 C/5 80 o 15 3 70 1— 1 H o •s. 60 55 o 2: « 50 o M 40 H O S 49 H 30 5= w S 20 w u 10 < w 5 ° N0NM0TILE MOTILE _L 2 4 FLOW RATE, gpm/sq ft FIGURE 5.12 EFFECT OF FLOW UPON PERCENT REMOVAL OF MOTILE AND N0NM0TILE NEMATODES 100, -3 50 o 1 40 o 55 g 30 3 w 20 a* 10 6 GPM/SQ FT- A ± ± j_ 10 20 30 40 PERCENT MOTILE WORMS IN INFLUENT FIGURE 5.13 EFFECT OF PERCENT MOTILITY UPON REMOVAL OF NONMOTILE WORMS 50 50 the filter should appear in the effluent. This number subtracted from the total worms out of the filter should then be the motile worms that passed through the filter,, It should be possible from this to compute the per- centage removal of motile worms , Computations of this sort produced extremely erratic results and were not successful,, Apparently,, the accu- racy of the data does not merit, such a refinement The conclusion that can be reached is thus only a qualitative ones there was some loss of motility by motile worms in passing through the filter o It is quite possible that the only motile worms which were removed by the filter are those which lost their motility while in the sand; those maintaining their motility were able to wriggle through the filter. But the data, although not contradict ing s are insufficient to support this theory, 5,6 Effect of Worm Size Influent and effluent samples from two filtration experiments at *+ gpm/sq ft were collected and the worm lengths measured using an optical micrometer. The frequency of occurrence of the different sizes in the Influent and effluent are shown by the histogram in Figure 5, 14 It is evident from this graph that the removal, of the worms was not dependent upon their sizes, A larger proportion of the larger worms appeared in the effluent than of the smaller ones. The. reason for this apparently anomalous behavior lies in the motility of the worms, Most of the larger worms were motile and were thus able to pass through the sand bed. The smaller worms appeared to be a nonmotile larval stage. Being nonrnotile s most of these worms were retained by the sand. co o O Pi M CQ s: D 55 180 170 160 150 140 130 - 120 - 110 100 90 80 70 60 50 40 30 20 10 3 "i 1 r i 1 1 1 1 r —6 INFLUENT Q ° |Q o. oo I !^_l L s=2L 51 20 OS w CO o 10 rs o t r — _L T - r t 1 1 r EFFLUENT J_ oy^i cx^o° P qo QISL _L .1 .2 .3 .4 .5 .6 .7 .8 .9 1.0 1.1 1.2 WORM LENGTH, mm FIGURE 5.14 NEMATODE SIZE DISTRIBUTION IN INFLUENT AND EFFLUENT 52 An experiment conducted with nonmotiie Triceph alo bus sp yielded a percentage of removal equal to 97, 9 Triceph alobus are about half the length of Diplogasteroides sp . , yet they were removed equally as well, again showing that worm size had little effect upon their removal by the filter, 5 7 Worm Distribution in Sand As mentioned in section 3,6, after several of the filtration experiments at 2 gpm/sq ft the sand column was taken apart by layers and examined for worm concentration per gm of sand. In Figure 5,15 the worm concentration, as a percent of that in the top layer (,1~,2 cm) of the sand, is plotted against depth. It is evident from this graph that the worm concentrations are highest in the upper layers of sand, dropping sharply thereafter. The worm concentrations decreased to less than 50 per- cent of the top layer concentration at only 3 cm depth. At 20 cm s or about one-third of the total depth, the worm concentrations found were never greater than 11 percent of the concentration in the top layer , one was less than 0,7 percent. There did not seem to be any factor which could consistently account for the distribution of the worm concentration through the sand bed. The lowest concentrations at intermediate depths resulted from an experiment with no motile worms. However, the highest concentrations at these depths came from an experiment with about 11 percent motility; between were the results from experiments with 38 percent motility, 12,7 percent motility, and 25,4 percent motility. The highest worm concentrations always occurred within the top 53 100 EXPERIMENT PERCENT MOTILITY 11 12.7 w u* o PC u >- 3 a. o 55 o M H < as 55 w o o s o o w o PC u .01 30 DEPTH, cm FIGURE 5.15 CHANGE IN WORM CONCENTRATION WITH SAND DEPTH 54 0„2 cm of the sand bed, with values ranging from 10 to 1.96 worms per gm, depending upon the worm concentration in the influent flow during filtra- tion „ Although varying with moisture content, the bulk density of the sand was approximately 2 gm/cu cm c This would mean that at least 400 worms per cu cm could be retained by the smaller sizes (upper layers due to classification during backwash) of the sand particles „ From the previous discussion of the removal characteristics of the filter, no evidence of overloading of the filter sand was found with the influent concentrations usedo The ultimate capacity of the sand for retaining nematodes was there- fore probably higher than 400 per cu cm To evaluate the importance of sand depth in nematode removal, it is necessary to find the distribution of the worms within the bed If a large proportion of the worms removed by the filter are found to be in the top few centimeters of the sand, it can be reasoned that increasing the total depth of the sand would add little to the removal capacity of the sando The converse of this is also true; if the worms are evenly distrib- uted through the sand bed then the removal capacity of the sand would be directly proportional to the sand depth „ In Figure 5„16 are plotted the results of five filter experiments at 2 gpm/sq ft„ The axes are the percentage of all the worms found in the filter bed below any particular depth, plotted against that depth, which is expressed as a percent of the total sand depth „ It can be seen that from 57 percent to 98 percent of the worms in the sand were found below the top one percent of the sand and that from 16 to 76 percent were found below the top 10 percent of the bed depth „ The differences between the curves plotted are directly attributable to the total worms removed by the I I I I I \ 55 o M Q t- 25 W to O H « • U m cu £ 3 .H O tu HId30 JO J.N3DH3d I I 1 56 filter. As the number of worms removed increased , the point in the filter below which 50 percent of the worms were found gradually moved downwards Or, putting it another way, at any particular depth 3 as the number of worms retained in the filter increased g so did the percentage of the total worms found below that depth. These data are plotted in Figure 5,1? for layers at 30 percent, 40 percent, and 50 percent of the sand depth „ It can thus be seen that worms are retained throughout the filter bed, al- though, from Figure 5„15 s the highest concentrations are found in the upper layers of sand. From this it can be concluded that an increase in sand depth should result in an increased worm removal capacity; but that this capacity will increase in less than direct proportion to the increase in sand depth. This is because a large number of the worms are retained at the surface of the sand, upon which an increase in sand depth has little effect. There is no reason to conclude, however, that increased depth will greatly increase the percentage of worms removed, except at high in- fluent worm concentrations when the increased capacity of the sand could manifest itself. Figure 5,17 has one interesting interpretation; if the straight lines of best fit are extended upward, they seem to indicate that at a total worm removal of approximately 30,000 worms, 100 percent of the worms would be found below any depth. This would mean the filter was overloaded and should be sufficient for at least an estimate of the "ultimate" capa- city of this filter bed. This number of worms removed on this filter (area = approx, 0,034 sq ft,) is equivalent to about 900 9 000 worms per square foot of surface area. Of course, from preceding discussions about 57 40,000 10,000 - g w s Pi o as H 1,000 100 □ 0.5 A 30% DEPTH, r O U0% DEPTH, r □ 50% DEPTH, r .99 .98 .98 J_ 1.0 10 100 PERCENT OF WORMS BELOW DEPTH FIGURE 5.17 PERCENT OF TOTAL WORMS BELOW DEPTH vs. TOTAL WORMS REMOVED 58 the removal of motile worms, the filter would have difficulty in removing this many worms if all were motile „ 5 8 Backwashing After several of the filter experiments the sand column was taken apart and examined for worms and then replaced carefully and backwashedo The sand was then re-examined to determine the effectiveness of the back- washing o This procedure is subject to several unavoidable errors It was impossible to replace the sand exactly as it had been 9 both because of the worms removed in the first set of samples and because of the destruction of the sand bed structure in obtaining these samples Also, since the backwash experiments were performed upon sands after different filtration experiments 9 the total worms retained in the sand for each was different; this could have affected the percentage removal by backwashing, although no consistent effect was apparent „ In any case the data obtained are shown in Table 5„lo There did not seem to be any particular advantage to backwashing at 50 percent expan- sion over 20 percent as far as nematode removal was concerned „ In this limited study, backwash times of from 2 to 8 minutes showed no consistent relationship to percentages of worms removed It does appear from this data j however , that under normal practice;, which usually falls in the ranges covered here (10, 12), worm removals of around 95 percent can reasonably be expected from backwashing „ As previously mentioned, the rather extreme measure of backwashing at 50 percent expansion for 30 minutes was used in these studies to assure thorough cleaning for experimental purposes „ The small, added benefit, 59 however, would certainly not recommend this practice on anything but a small pilot plant „ More thorough study of this aspect is needed;, with better experimental techniques 9 before any more definite conclusions can be reached „ TABLE 5.1 PERCENT WORMS REMOVED BY BACKWASHING Time (min) Percent Expansion 50 20 2 95„3 5 69 o 7 98 8 96ol 97 „ 9 8 95o6 30 100 100 60 60 CONCLUSIONS lo Removal of nonmotile nematodes was independent of flow rate or concen- tration 2 Removal of nonmotile worms was about 98 percent „ 3, The percentage of nematodes removed was constant during the 8-hour filtration periods. 4. About 25 percent of the motile worms were removed, independent of flow rate, 5o The percentage of worms removed decreased with increased percentage of motile worms in the influent, independent of flow,, 6 Percent worm removal was independent of concentration at the flow rates, concentrations, and lengths of filtration period studied„ 7„ There was some loss of motility by motile worms passing through the filter; this percentage increased with increased flow rate 8 Worm size had little effect on worm removal „ 9o Worms were removed throughout the sand bed with highest retention in the uppermost layers 10 o Increased sand depth will increase removal capacity, although in less than direct proportion to the increase in sand depth 11 The concentration of worms in the lower layers of sand increased with increased worm removal by the filter „ 61 7. SUGGESTIONS FOR FUTURE RESEARCH lo Quantitative evaluation of the percent removal of motile nematodes, perhaps by pretreatment of the influent to obtain only motile worms (100 percent motility) „ 2 Quantitative evaluation of the motility loss by worms in passing through the filter bed„ 3„ Investigation of various means of reducing motility of influent worms, thus increasing their removal,, 4„ Investigation of the possibility of using Zoopagales fungus , a "nematode trapper," to entangle the nematodes, thus increasing their removal by the filter „ 5„ Investigation of effect of turbidity in raw water upon nematode removal by rapid sand filtration, 6 Investigation of the effects of pretreatment on nematode removal by rapid sand filtration. a c Coagulation and sedimentation bo "Coagulant aids" Co Polyelectrolytes 7. Investigation of the effect of sand characteristics upon nematode removal by rapid sand filtration,, 8o Investigation of the removal of different species of nematodes by rapid sand filters „ Also of mixed populations , 9o Investigation of the effect of temperature upon nematode removal by rapid sand filters . 10 o Investigation of the effect of high concentrations of nematodes and 62 long filtration times upon their removal by rapid sand filtration, a Sand overloading and "breakthrough" b Nematode breeding within the filter 11 o Further evaluation of the effectiveness of backwashing in cleaning the sando 63 REFERENCES lo Cobb, No A., "Filter Bed Nemas ; Nematodes of the Slow Sand Filter Beds of American Cities " Contributions to Science Hematology , 7_, 189 (1918). 2o Chang, S„ L , Austin, J H., Poston, H W„, and Woodward, R L , "Occurrence of a Nematode Worm in a City Water Supply . " J Am Water Works Assoc , 51, 671 (1959). 3o Chang, S. L., Woodward, R, L. , and Kabler, P„ W , "Survey of Free- Living Nematodes and Amebas in Municipal Supplies." J. Am. Water Works Assoc o , 52 , 613 (1960). 4. Kelly, S. N. , "Infestation of the Norwich, England, Water System." J. Am. Water Works Assoc , 47, 330 (1953). 5. Chang, S. L,, "Survival and Protection Against Chlorination of Human Enteric Pathogens in Free-Living Nematodes Isolated from Water Supplies." Am. J, of Trop. Med. Hyg. , 9_, 136 (I960). 6. Engelbrecht, R. S„, Dick, R. I., and Matteson, M. R., "Factors Influenc- ing Free-Living Nematodes in Water Supplies." Civil Engineering Studies , Sanitary Engineering Series No. 18, Department of Civil Engi- neering, University of Illinois, Urbana, Illinois (September 1963). 7. Hirschmann, H., "Gross Morphology of Nematodes." Nematology . Edited by J, N. Sasser and W„ R. Jenkins, The University of North Carolina Press, Chapel Hill, 480 pp. (I960). 8o Ward, H. B„, Whipple, G. C, Fresh Water Biology . Second Edition, Edited by W. T 8 Edmundson, John Wiley and Sons, Inc., New York 1248 pp. (1959). 9. Chaudhuri, N., "Occurrence and Controlled Environmental Studies of Nematodes in Surface Waters ." Ph.D. Thesis, University of Illinois (1964), 10 o Water Treatment Plant Design . ASCE - Manual of Engineering Practice - No. 19 (1940). 11. Riehl, M. L„, Water Supply and Treatment . Ninth Edition, National Lime Association, Washington 5, D. C, 223 pp. (1962). 12. "Backwashing Sand Filters ." Water Works and Wastes Engineering , 80 (February 1964). 13. George, M. G., and Kaushik, N„ K„ , "Infestation of Surface Water Supplies by Nematodes." Environmental Health, 6, 229 (1964). 64 14. Robeck, G 5 G„ , Woodward, R L. , "Pilot Plants for Water Treatment Research," ASCE Proceedings,, Second Edition (July 1959) 15 Symons s J, M. , "Simple Continuous-Flow, Low and Variable Rate Pump." J. Water Pol lution Contro l Federation 35, 1480 (1963). 16. Chaudhuri, N», Siddiqi, R. H„, and Engelbrecht s R. S., "Source and Persistence of Nematodes in Surface Waters." J. Am„ Water Works Assoc , 56, 73 (1964). 17. Chaudhuri, N., Dick, R, I., and Engelbrecht, R. S., "Staining of Free-Living Nematodes by Eosin-Y Dye." (in preparation) 18. Baliga, K. Y„ , "Benthic Sampling, Analysis, and Ecological Studies of Nematodes." M„ S„ Thesis, University of Illinois (1964). 19. Volk, W. , Applied Statistics for Engineers . McGraw-Hill Book Company, Inc., New York, 354 pp„ (1958). 20. McCracken, D. D., Dorn, W. S., Numerical Methods and Fortran Program- ming . John Wiley and Sons, Inc., New York, 457 pp. (1964). 65 APPENDIX A WATER DATA 56 >H P « LM < S CD 1 JEI to 3 \, CO E G< < bO H < CN Q co co CD CD CD CD st r- CM [■>• CD CD CO H CD CO O LT> CM o o l> rH cd r-» o o CO CD CD CD CO st H co lil CD J- LO CM CM CM CO CM CO CO rH LO CO CM t^- O CO CO cm cn CO LO o O ° O CM CM CM LO LO rH CM O CM rH LO LO o H o O H- * cd CD o H O -H o O « o o CO LO LO l> CD cn St COCDcDCMCOzfrHCMCM d-d-oij-inoooHri LncOiHOcocDto iLO * d - H co co Hcod-jod-como J-LOCOOt>COCOCOLO CDCOC^LOCOCOiHrHCD J- CO CM rH CM rH oimjCNood->>o OOrHt-^O^-LOCOCDCM r^or^or>CMt>-CMLO LOCMCOCOrHHCM CM o o cn cn rH rH H H rH H CD CD oo o co o t-^ o r-~ CO CO CD CO CMLOcOCDrHCOCOLOCO LOr^c-~CDLO^-LOCMCM c~-LOrHr-coj-cncMr^ CO CO CM CM CMcncor-cocoLnLOO d-omcococohCNcn cococococod->jco CO CM CM CM co co t> r- CM CN O O CN O CM O CM CM H CO r> CD St r-- co co i> zt > I CO CO CO CO CO CM CO CO CO st 00 zf rH CD CO LO r- o cn co cn co cd in lo co f- CO CO rH St CM CD CD CN CM CD CN CN CN st O LO CO CM H/ O CO J" t-» H co LO LO CO O O O O o CO CO rH rH CO CD St st CM LO rH CN o CO st LO LO o o O o o o CO cn CD st CD cn LO st LO LO LO H/ CM CM LO LO CD CD CO rH O rH o O o O o rH CM CM CD CD CD cn LO LO LO LO CO LO CO o co t-- o CD H/ o I> CO CD LO rH CO CM CD LO CO CD co C- co CD o CO CM CD -i- CM CM o CO o o o o o CD rH l> CM H CO rH LO o o o o o CD CD rH CO H CM CN CM CM CM CN rH cn co LO CO CD CO CD CM H/ CN r> H H O CD CN C- CO CM CD CM rH 00 CO en cn r>- co CO r- cn o co o CM r>- CD 1> rH o on H r^ O o C- o o o o o o rH CO CO H/ st CD t> co J" o o o o CD LO rH CO CM co CD St rH CM CO H rH o o CD O h,- CM CM LO CD rH CM 4-» -H rH rH rH r~- CM co CD H P G CD st P G (0 c d T3 > 3 G M o > -a CD O E H cp rH u T3 > CD G Mh CD E 3^^\ •H P •H 1— 1 o O 04 a* CO G G c G H 3 H P E cr; E •H •H •H M \ o "^ C CO G CO CD CO -d Sh E CO to M E O E K E E 5 E E E CD rH CD rH to o to o CD to o o U U rH G rH 4-> o P O E fe E rS > E & E o tu O O O r-^ •H 3: •H & rH rH O U 0) s= rH 13 DB •H rS P H rH O 0) p o 0) E o CD P4 ■H P Mh ^""s ■^ rH G ST -H CD E3 H CD p > > > O o o o O O O O O - P PC Mh < s cr >■; CO o "■»>, 00 e a, < bO E-> < d « en co en CN o CO O CO CM Cn CM co H CD in 1 rH cn co H d" H CD H cn H in O in in cn in cn r> CO CO o o O o u o o o d m CO CM d" CO rH H cn o o a o o CO H CO H CM d" CM in rH O en H cn CO 00 CM CO rH cn in CM CO d- CO rH CM CO CO d" CO CM CD CM H rH r> rH in en CM 03 i_n cn rH co H CM CD d CM r- r- o CD f* cn t-~ CO CD (XI o o o o o [> co d CO r- H co in co o o o o o r~~ cn 00 ■H co CO CD H r^ H rH CM cn .d H CO CD CD 00 in d- CM CD CD cn H in CO d CO d en cn J- o CM d CO CD in o CO cn CO en CM CM CD cn CD CD o in cn d m CD in o rH in rH in o o m o u o o o o CD rH in CO CO o CM r- in a o a o 3 cn CO H en CD H CO CD d in m CO CM d co d CO CO CM CM CO H CM rH d- H CD d- CD CD m uo co O o CM CM CO CO en co CD CO r> r> cn CD C- r^ CM CM ° cn o o a O o O H o H o O o o o O a o O co CO CO H en d- CM rH CM H CM H CM rH cn cn cn cn d CO CM CD o in o CM CO d CM H CM o r> CO cd co CD CO CM c-~ 00 H CD d r-- C- d- d cn in CO CO CO d o t> o o a a o O CO r- o H co o H CO o c o o o CO 00 CO •H CD H d d CO rH co CM d CD CD r> CD * in co d- co CM CO H CO in c-» CM J- CO CM cn CO in co cd O CD o o o o o o CO CO rH CD CO f- CO iH cn cn co co H rH CD cn co co H rH d- cn d" CO iH O H cn d co E-i CO U E- P P cDMn CO w E co CX co bO O O fd rH CD CD O O rH d- O CO CD H o cm r> CM rH H CD 00 C- d o O o O CO CM co d- d- «H O) b co co in cm co rH CM in j- cn cm CD in CD CM CM CO CO CM CM o p. a) p •H C rH IH \ CO U E CD P. •P o •H Es rH ■"V CD CO -H P. cd ■p P -H 3 rH o CO £ u o V. (1) CO rH e -h 5 p o o CD CD CD CD hO bO bO bO td to nj rj P P P P CD CD CD CD > > > > < < < < OOOCDCOrHrHCMCD d-d-ot^cod-cDOin OCDrHCOt-rHrHCMCD d CO rH rH CM H OCOCMJ-drHcDJ" OHM^HlDHO Hd-C-CDDOCMLn m CM CM CN rH rH CM O O CD CD d" CD CD rH rH d" cn o CD H O H co CD CD CD OOdCODOOOCDCDCO COcDCNCMCOCninCMCO CMCDcDCDd-d-COCNiH cd in in in cocoomcDDoor^rH iDmrHOCOCDLOrHd dCOCDCNLOCDCMCOCD If) =t r-\ J- CO o C co H 6 u CO O e es CD CD WHO e -h e f< p a o o o rs s: s p H O p o CO E o CD CO rH p o CO E U T3 O CD 2 > o e CD h c*; CD > X) o CD p CO E c o 53 rH rH rd CO p p o o H E- O Pd E CD CO « E C-l CO o E & O CD 3: rH •H CD P rH O •H E P c o o rH cd p o in m cm co co CO rH CO CO CD LO CM CD CM CO CM O CO CD CM in J- rH O f» CO CO CD rH d- o cn in id in co r> in in d t> co co rH d- p p CD 3 3 O rH rH E LP Mh CD C Uh PC M w T3 CD > o E CD PC (D 10 > E o ^ E O CD IS PC CD CO rH E -H U P o o co c C E 'H -H p. O CO CO SEE r^ r^ CD O O P CD CD O H H E *H -H C P P o o o s s s p p p p p C C G C CD CD CD CD CD o o o o o S^ P. ^ ?J r^ CD CD CD CD CD 0-, 0-, 0-, 0-, Oh o 00 d- rH r- co r- CO CO o r> o o o o o CO rH CM o rH CD d- d- d- CN CD d CD d d co CO o CO o O o o o CO CD CO D CO CD d- o CO d- o CO d- C O •H P rd P. P H •H Uh Mh O CD E CD P. P H O CD E rd co CD P. p H O CD E rd co CD P P rH O CD E rd co 68 >-< 4-> £* Lp < S a 1 s to 3 \ co s a < bO E- < to Q cm co co H r^ co o o o cn r~- m r-\ o D O o O O Q ■H in cn o CO CO CT> CO J- cn o CD o CO CO O CO r- cn to o o o CO IT) CO CO r- o o o o o o i> CO CN CO CN -H o o o to to cn cm cn CO co c-- H rH rH to co o r- 01 in to CM CM 10 H > H n -H cn rH H CO co o CM to -H i> cn CN o o o o o cn to CN CM co d- H r^c^omLOrHCMCMo mmor^coj-cocMto CMtHHt>r-Od-^fO O rH CD ,-H H CO CO CJ) rH in co co CO co m cn rH co to it r- o co CM r- to i> cn cm H CM r-| H j- cm r-^ m o cn r^ H r- r^ i> o ho o rH rH CO CO CO =t 3- m cn t> cm to co co co MNjmHi-tNrlO COCnj-rHCnCMCMOCM cncoincococntooto rH cm cn co rH H H CO H r- m to co j- CN CO CO rH H CO CO CO «H J" j- o in cn r> co co m j- cm c- co co H zr cm cn co o o CM rH cn O CM c- 3- t> cm j- en O J- H cn rH m CO o o o O o O Lf) CM d- CO CD CO tO CM CO o en d- o o to to o o o a e a t*- CO rH m CO O ^ ^ j- 1> f~ CO 00 cn cn o o co co cnoairHOrHcooco 3- j- o- =r cninj-inzfrHj-rHco CMOCMCMOCMOOO oocMcocnHcocnHco J- zj- co co COcOOHlOtQCNCMO CDrHr— ocotocncorH coiHCMintococoj-J- inCMCOCOrHrHrH H rH CO O CM i> in cm co cn co co cm zf co J- o CO zt d- [>■ in co Zf CO ZT U3 cn m CO r-» o r- CM Cn O d- i> to o o CO H CO rH ID CN CO in CO cn CO CN in CO o a o o o r- CO CO CO CD r- CO CN o o o o © o r*> IT) CM in co J" rH in CM CO H co r-« to CD en CO CN 1> r> CO 3 CO rH rH zr o r-. =h o CO zl- o CO zt- o CO J- cn cn in a o ° o o zr c^. r- 3- cn cn CO CM CM CN CO O r> o o o in O CT> CN zf H zr CO J" CO H o m .zr p P id c c p 3 3 M O cy > T3 CD E tp rH LP !h U TJ o > 01 C LH a> CD E r~\ •H •p ■H M o O Qi CD to C C C c rH 3 rH P E « E •H •H •H rH V, O ^^ C W 3 CO a) to -d rH E IZ) W M E o E P4 E CD to O to to Sh E f-, E s^ & T3 fc > E ?s E E a; H a) ?H to o CO o a; to O O S-. H rH C p o p E 2; E ^ > E 13 E o CD O o *-~-s •H 5 •H 12 u & o t. CD ~£ H [2 S »H p H H o a) P CD E CD &, •H P 4h *-> *\ a) ^^ (D C ^3 H d ^2 H cu ^S H CD P CD CD rd P M H (0 rH rH •H o •H a-: •H to H H rH Sh tT'Lp £ •H e •H a) P ■1' P cu P E •H E •H •H p CO w U •P s^ P W H a m rH o (0 -i o fi ^ C P P H *^ O o o E •^ E E •H E E •H E o o O •H E to £S ^-' 12 s r^ p C f4 p c Sh P C 13 ;s a gj S P- a. to O o O o O o O o bO o (1) a) 0) > > > O o O o O o O o O CD (1) CD a< CD •H U< tH < < < < H E-< ^ H E- H E-i Eh H Ph Pu Ph Ph Ph E- CD rH P H O CD E rd co CD U P H O CD E rd CO 69 APPENDIX B CORRELATION COEFFICIENTS 70 KEY 1 flow rate (gpm/sq ft) 2 head loss (feet) 3 average worms per liter in 4 average motile worms per liter in 5 average worms per liter out 6 average motile worms per liter out 7 total worms in 8 total motile worms in 9 total nonmotile worms in 10 total worms out 11 total motile worms out 12 total nonmotile worms out 13 total worms removed 14 total motile worms removed 15 total nonmotile worms removed 16 percent worms removed 17 percent motile worms removed 18 percent nonmotile worms removed 19 percent motile worms in influent 20 percent motile worms in effluent 71 oinojiDHcnohcncMoo CO W o M +J Cm w a 1 o w o E E-i CN - o r- o r- C"~ o C^- en LO co CM o H CM 00 H o o co CO CO CN d CN CO CO 3- o d CM OO o H LO LO o d OO CN co 3- LO CO d H r- o CO l> CO CN CO LO OO co LQ LO LO CM H CN o LO CD d CO CO r- CO LO co r> o LO co r-» I I ! OCNC0C0C0C000HC0C0r-cr>C0Hr>-CN ohcoldocnlolooolocococnlocoh ooooocNcocncod-rocoHLOcDcocoLn ocncDCDLOOoc7>cr)drLnd-coOL--r-LO H III OOCMLDOzJ-cOHCOCnLDCOH^uDCOCO Ot>CT)CDrHiHLDd-CMLDr^OLOOu-)LDO OHC"-CNCDOOOt>LOOOCOLDLDCT>CNLOCO ocj)cr)LDcnd-cj>cr>r^coc~~oot^OLDcor-~ oo I OLCCNOominro-CNcncoOLOOcoiDHr-- onoodCDHCD co cm H co cd co co co r-~ orc-coro~cnd-r-cncot--codHcnf-Hm ocDf~-tDcnt>cDr~r^r-~cj>LOCDCNOCNCM ) l l CM o o o o r-COCDLDLDd-LDCO OOOLOfO-CDCOCOH oo oo cm r~- 00 CO CO CM 00 O ID O0 O O CO in oo oo o d cd LOCNOOCNLOCNCOHCNH dcodcod-oomcDHOi— I CO LO CM CD f- CO 00 LO H lo d d O H I OCOLDCOCNCNCNHOOr^COHCOOOCOOr--CnCDH OLOcocNdocjiLor-CNrHd-CMCMOomcNcnr^cM ocoooina>aiLDcodHCNCDd-codcocDLDCOLO od-coHHHcoodHHOioommomd-oo H I I I rHCNCodmLDr~cocnOrH III! CN00J-LnuDC~-C0CT>O HHHHHHHr-ICN r- o o o o O CM o H O LO O CO O CO CD o m cm O CM O o c-- in O H CN CN oco m^ O CO 00 CO O O CN d o H H d CO o H H r> LO o d- r- d LO o o 00 CO LO H I o c co r- CM X> o H co c^ CM LO o 00 H d O LO o o H H H H rH o o o o rH O CD LO CD t> d CO CM CO d t> CO LO CO CD t> O CN p- to- OLOJ-J-LDC30COCO Or- I CO O H O LD LD OOOOLOCOCNLOCD OJ-OOOOrHOrH o H co CM co LO CD o o o CO 'Xi CM d d CO CO LO o co co H f» LO LO co o o LO LO LO' t^ o J> LO LO I I I o d o co CO CO CO co LO H o LO d d r^ LO H d CO d o CD CO o LO CM C- CO LD H o [*■ 3- ID d r» o LO !> [-- CN CO d H H H I I lo co r- H H H CO CD O H H CN 72 o cd r- o o r^ OlinHtDCOHHCN LO ID CD 00 CM LO CO Od-HMWOJNCOOMD ocncDcocDcocoocor^LO CNr"-LncOCOCMCOLOOLnCD LOiHrHCOLOOt--OOCDCD co oo o C"» m co c-~ H (N O Ol tN O) m rH I I I 00 (D d- d- zt in in m I I o o o o o o O LO O LO O 00 to H S3 w n o M +J P-. M-i W cD O CO O \ E O M M H z}- < W OS o o o CD CD LO CD CO CM m CD o c-~ CD CD CO LO H c-» O o CO Zt" CD zf t> c«- CO J- o CM CD CD C- co CD =j- co zl- lO rH I O LO O CO rH O H CM J" C~- O CO H CD 00 OH Ol (NH CD CO H CO H CO CM LO 00 CD O f» CO t- I CO CO r-~ j- J-lDOh . i-HCMmcOCMCMLOCOCM CO00HI>CMJ-CMHCM CO CO zj" O rH cn co h J m H o o o o O -COzt-CM oia)HoicNoio)>moij-coo I in cd co r- C0 CD CO l> 1 I I co co in > n j- H m J- CD CM a> co co cd co CM rH CD in cd t*» . O CD CO .— llDCOCMCMincOCMCOd- rHCDCMiHOOCOrHr-CM-j-CM co H oo £-~ CM CO zf oo It- CD CM ! 1 I CM oincojHOOoicoinoiooioiHa-CMhOi oinj-coj-inLninincor^ini>fHLniHcooj-CMo i i i i i I I I I i I i I OCOCDCMinCDCOCOHCDOCOO OCOCOCMOOinCDr-OCOCOCOCM or^CMcorHcococMcocozj-zj-iHLn CDCMJ-CMCDCslJ-zfCMincOCMCOrH CO CO CM CD Zf H cd cd in n- h O CM o o I li-(CMcozriniDi>cocn O iH CM CO H H H H i g zj- m co r^ co cd o rH rH H rH H H CM t> o rH LO CO O CD CM rH ID C O H CO m r- o rt cc zt rH CD CD CD f» CD CO LO zf zf o o CO CD co zj- CD CM LO LO CM co rH LO co in o H CD CM CD CD f- co CD zf CO o r- co c-> CM o zr CO o CO rH o CD O o CO r^ rH o c o O CM H 1 O LO zt" LO CM CD CO CD O CO o tn CO CO m o r— CO CD Zt" o o co co cn o CM CD CD r- O J" CO CN CO ^ O C- rH CD CO CM o zt- r- zt- t-» m o o zj- r- cd co I I l O O zt CD H C- CM O CM CO H O rH 00 o r- co o t~- m rH o zj- oo cm t- co co o CO LO H CM O rH o o o t> CO cm in CO H o r-4 CD o r-^ H CO H o zt- CD r- zf r- LO LO II 3 o co co r> o ld O t> CD O rH CO O!>O00CDCDCMC0 ocMcoojcOrHr-in 00 Zf CM CO zt- H I I i i CDrHlOOzl-CMzj-mCD COCDLDCDLOLOCMrHI> CMLOmCMCOrHCDCOzl- ococDzfcooir-cot-- II » i ^HCMcozj-mLor-cocDO rHrHHrHrHrHHrHCM 73 CO u M CJ M -P Ln Mh o w o o hfl H in < W o o Icn co OCOOOCMrHCNOOCnjd- OlT~-^H.HOd-OrHQ0J-CN ot--cDomcn:3-cocMCM[-~ ooia)OJHH>nja)co o'oooooooooa r-\ 9 II I OCNCTiJ-OrHOO CD CO OOOd-CNlDCOniDP) OrHcoCMcDr--cocT>coco oj-nmmjoioon o o o o O CO J" ["- o o =t H O CO CM j- O CO CO h (N o r- cm co cn co co CT> cn H H O CM o co i- io cm r- zt-cncnr~-CNCMCMLOcocococo co OHr-cocDHcoo OMOd-NCOd-d- OlDOIlDHJ-COCO O[^-C0t^C~-C^t-J- H oHownoioi(0> OJ-t^coc^cncooj- oococnc-~ct>cocoh- ohoinoioibdo co H cd r~- en co co cd cd en in o CD CM a i in cm c^ o J" f- f- r-~ •-{ en t— m cm co co cm m co co co co m en H •-{ co j- co r- co I I I OC0CDC01>CDiHJ-CMC0C0OOCTlC0CT> ocooiriHtJicococomomcoMDW oc-~.3-cocncnt~-Lnc-«zi-r^inco.d-coc^ ooif-aicoo)0)coHHH>n^coco CO CM ococoj-cocncnj-Hcocnmooc^coo cnmcMCMCsicocor-iJ- r~- o o en o r- cm cm i-H r- ocMr-cocncoiHr- ococomcncococo Ocncnt-»cncocDC7> J- c-~ co r- co co co co o o o o COcOCMHCDCOJ-r~!>- cni-HcDt--cococMin.3- I I MO 4- O W mn d- co J cocMCJicocor--CM^i-r^cMcOE^CNtncor~ t~-c^-cnr^co!>i>r-c^^-c--CMrHOd-c^ o rH co co CO =J- CD :* J" en o o J- CM en r- CO LO o l> o CO CM i_n H CM CM CD LO rH o H O o o -H O rH o o aiCMCMCMCncOCOCMCMCM CO CO o r> en in co co H zf H H co o O H H I I I I I I I I I I OH-r-cDd-CMj-incococncMHt>co^-H-cocnrH ocoinHcorHcDcnrHmcDcooco^-coitmcoco ocncorHCNzfcDrHiHcoit^rmr>r^cotncot>-m OOmocDr-Lnr^COCOC^CMCMOCMOCOCMCMCO H I II 1 I 8 1 I I 1 ii 8 g CMcoj-mcor-cocnorHCMcozt-mcoi>cocno H.HH'HHr-lr-lr-lr-HrHCM o o o o o o O CM o co O CO O O CD O zf o O O H o co m o en CD CO o r- rH H o CO CO CD o o CM O H o zt- c-- o r- o r> cm 3- co o co cm m m o rH en en c~~ o o o o rH cm en cn rH CO CO rH co co in en cn rH in rH CM o o d- rH CO cn CO o CM r-~ J- LO zt cn o CM c-~ H f» o r- o J- H LO H CO 3- o =t Zf zT co rH cn cn o LO J" l> LO CO lo CM o —1 cn O CO o .=1- CO o LO cn co H LO H CO o r- O H O O f- O CO CO CM H in zj- r- h r^ co co co in co r^ co co o in cm in in zf iH m j^- co I 8 OLOCMLOCOCOCOCncOCO OH-cocom.H-rHococM CO ^ 3- C^ H O co rH co CMcozi-LOcor--cocno rHHr-trHrHrHf-JrHCM 74 LO CM It H ID H W CO m cd r- c-- cd co h m o CO CO 00 CM en r- ID CM CO CM CM cm in o co r-- c— m o o o o j- co d- CD CD in in id co CO CM it co cm iH in in o co o o rH zt CO CO CO o in in en co cdcocDi-iocmoo tr- ee w I— I o (—1 u^ co H O O M C_) Ui S H O rH H < < g s o o OltHCOrHCOCOCOC— oitcnc-cncoHcoco OOCDCOCDinCDOO ococDCDr-rHinrHr- MDCO J t O in CM CM CO O ID o in co r- ocDr~c--i>itmcMOfHcoomcM ocDrHCMiDcococoinitJ-cocno OCMCOlOCDltinOCNcOCOCOCMlD oiDcniDLniDcomcor— ioocmco o o o o zt m co m H CD It CO CO CO > > It CO H CO CO CO O It CD CD CM in co H m H H r-» co m CD H m h O C- H It It CO CD f- O ID (D CO H ID in co f- o o o o H co co co m o rH It r» CM CO CMor-tomcDcoinio incOCMCDlDCMincOCOOCD COOCOitCOCOH r-coo:tot-ocot--r- O CD O CD CO O CD CD It CO H inc0CMttC0(0^(DtOtm rHCDCOLOrHr- CDOlti— IOCDCO r-cDcocoj-^ir-cOiHcoo cocor-omor-ocococo CD CO CO O CO f» O C-- CD r-1 O CO O CO It H It o o H co CD CM CO m (D to t>COCOCMC0C0C0C0cOC0mc0CO zfinitcDCMiDcocMcoiniHino o o J- CO O rH CM o r- CO o to CM c-» CO in CO ID d- It cn m CO o CO H o o CD CD CM CD H l> H CO r» CM co CO o H CO H H o Lf) CO d" CO f» it CM :* en j- CD o -ot>cDomcocorHcoj-oooco ocooincocooor-comitpHmoitCMrHCDtn OCOrHitCOOOt>tOOcOHCOCOCMCOOOrHr-OCO OCDCOOOOJ-COltCOCOCOCOrrCOrHCMnHCMH i II JrHCMcoJ-incDr— cocdohcmco it mcotr-cocDO rHrHrHrHHrHiHirHrHrHCM I CD IH co o o o o O CD O rH O CD O CO o cm in O CD O O rH CM O CO CD O CO 00 H O CD CM rH O It CO rH O H O CO O O O O H I O It H rH CO O CO CM CO CD O O ID CM ID O rH CD CD f» o o o o H o LO ro CM CO CM o CM It CO It in o CD r> CM en CD o CO o d- CM O O CO co r- co J" o =t o C"» CD co CO o t> CM H LO o =f o r~\ in rt J- LD CO o CO CM cn I'M CD O ■t o CD LD CD LD LD in LO o in CD O CO f» CO CO o CM CD CO o co CM o CMOLDitOCOltit CMrHf-COt>OCOr- I COCOt>COCOCDCOCO CM(DrHr-rHcDCD(D o o CD CM H in r- rH m O o o CO CD o o r-» Il- o in o CN — 1 LO CO CO o CD !> H o r> rH =J" o LD H d- p~ f- .CNfOtmiDhcoroo IrHrHHrHrHHrHrHHCM 75 APPENDIX C SAND DATA 76 SAND DATA SUMMARY Test Depth to Top Number of Worms Total Worms of Sample per gram in Removed Layer Layer During Filter (cm) Run 56 101.9 o3 26.6 7,5 16 1„76 9.71 7,059 32 .96 48 1.17 63 .3 58 10 .0 o2 19.0 6.6 14.4 1.12 0.87 10,791 30.4 0.36 46.4 0.42 61.8 0.078 60 29„2 .2 28.2 6.7 15.2 6.67 6.34 14,373 31.2 6.9 47.2 8.35 62.2 3.6 68 196 .1 110 3.4 11.9 9.1 1.5 2,152 27.9 1.15 60 70 12.55 .1 19.5 3.9 12.1 2.1 220 28.1 60.4 .3