SOIL AND GRASS FILTRATION OF DOMESTIC SEWAGE by Gordon Stanley Lehman

SOIL AND GRASS FILTRATION OF DOMESTIC SEWAGE by Gordon Stanley Lehman

SOIL AND GRASS FILTRATION OF DOMESTIC SEWAGE

EFFLUENT FOR THE REMOVAL OF TRACE ELEMENTS by

Gordon Stanley Lehman

A Dissertation Submitted to the Faculty of the

DEPARTMENT OF WATERSHED MANAGEMENT

In Partial Fulfillment of the Requirements

For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

1968

THE

UNIVERSITY OF ARIZONA

GRADUATE COLLEGE

I hereby recommend that this dissertation prepared under my direction by GORDON STANLEY LEHMAN entitled SOIL AND GRASS FILTRATION OF DOMESTIC SEWAGE

EFFLUENT FOR THE REMOVAL OF TRACE ELEMENTS be accepted as fulfilling the dissertation requirement of the degree of

DOCTOR OF PHILOSOPHY

DisserEation Director

Date

After inspection of the final copy of the dissertation, the following members of the Final Examination Committee concur in its approval and recommend its acceptance:

/

//t

/2

7

747

/

7

/

7

*This approval and acceptance is contingent on the candidatets adequate performance and defense of this dissertation at the final oral examination.

The inclusion of this sheet bound into the library copy of the dissertation is evidence of satisfactory performance at the final examination.

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at The University of Arizona and is deposited in the University Library to be made available to borrowers under the rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgment of source is made.

Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate

College when in his judgment the proposed use of the material is in the interests of scholarship.

In all other instances, however, permission must be obtained from the author.

SIGNED:

4

J

ACKNOWLEDGMENTS

The author wishes to express his appreciation and thanks to

Dr. L. G. Wilson for his guidance and encouragement throughout the course of this study and in the preparation of this dissertation.

Thanks is also extended to Drs. T. H. McIntosh, D. D. Evans,

J. H. Ehrenreich and J. L. Thames for their guidance and critical review of this manuscript.

The author also wishes to acknowledge the staff members and lab technicians of the Department of Agricultural Chemistry and Soils who conducted the atomic absorption analysis.

Thanks also to the members of the Water Resources Research Center who aided in the construction and operation of the facilities employed in this study.

Special thanks to the staff members of the Sanitary District

No. 1 of Pima County for their co-operation throughout this study.

The author wishes to express his appreciation to the Sanitary

District for providing the sewage effluent and a research area for grass filtration studies and to the Water Resources Research Center for providing other research and laboratory facilities used in this experiment.

The work upon which this dissertation 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.

111

TABLE OF CONTENTS

LIST OF TABLES

LIST OF ILLUSTRATIONS

ABSTRACT

INTRODUCTION

LITERATURE REVIEW

General Waste Water Reclamation Characteristics

Trace Element Filtration

Plant and Organism Response to Trace Metals

Techniques for Trace Metal Detection

MATERIALS AND METHODS

Materials

Sewage Effluent Supply

Grass Plot Facility

Lysimeter Construction

Methods

Effluent Application Rates

Sampling

Chemical Analyses

Coliforrn Enumeration

Physical Measurements

Data Analysis

RESULTS AND DISCUSSION

Effluent Application Rates, Losses and Infiltration

Rates

Physical Environment of Lysimeter Columns

Trace Metal Filtration

Iron

Manganese

Chromium

Nickel

Copper

Zinc

Lead iv

13

13

14

18

24

24

25

27

29

30

30

33

Page vi vii ix

1

4

5

8

9

10

13

61

61

64

70

75

33

42

46

49

54

TABLE OF CONTENTS--Continued

Cadmium

Cobalt

Strontium

Trace Element Removal at Grass Plot Site

Nitrogen Compounds

Coliform Organisms

Chemical Oxygen Demand

Recommended Treatment for Soil Filtration of Sewage

Effluent

SUMMARY AND CONCLUSIONS

REFERENCES CITED

V

Page

97

100

105

79

81

81

85

90

94

96

LIST OF TABLES

Table

Soil Textural Analysis of Distinct Horizons for Soil in situ at Crass Plot Site

Textural Composition and Bulk Density of Soil Material

Packed in Soil Columns

Weekly Volumes of Effluent Applied, Estimated Losses by

Evaporation and Infiltrated Water for the Lysimeter

Columns

Cumulative Trace Element Applications to the Four

Lysimeter Treatments

Initial and Final Total Sodium Acetate-extractable

Trace Elements in Lysimeter Soils

Total Quantities of Trace Elements Removed in Grass

Clippings from the Lysimeter Columns

Concentrations of Trace Elements Detected in Filtrate

Samples at Grass Plot Site

Concentrations of Trace Elements Detected in 9.1 and

15.2 Meter Well Samples at Grass Plot Site

Page

16

22

39

47

48

50

87

87 vi

LIST OF ILLUSTRATIONS

Figure

Lysimeter Construction Showing Arrangement of Soil

Columns, Effluent Distribution System and Instrumentation

Face View of Inside of Lysimeter Collection Chamber and Typical Section of Soil Columns

Total Volume of Effluent Applied Daily to Continuously

Flooded Columns

Total Volume of Effluent Applied Each Day of Irrigation to C-2F5D Columns

Total Volume of Effluent Applied Each Day of Irrigation to 1F1D Columns

Total Volume of Effluent Applied Each Day of Irrigation to 1F1D-1F3D Columns

Quantity of Iron in Applied Effluent and in Filtrate

Samples During Lysimeter Treatments

Quantity of Manganese in Applied Effluent and in Filtrate

Samples During Lysimeter Treatments

Quantity of Nickel in Applied Effluent and in Filtrate

Samples During Lysimeter Treatments

Quantity of Copper in Applied Effluent and in Filtrate

Samples During Lysimeter Treatments

Quantity of Zinc in Applied Effluent and in Filtrate

Samples During Lysimeter Treatments

Quantity of Lead in Applied Effluent and in Filtrate

Samples During Lysimeter Treatments

Quantity of Cadmium in Applied Effluent and in Filtrate

Samples During Lysimeter Treatments

Page

19

20

34

35

36

37

52

56

62

65

71

76

80 vii

LIST OF ILLUSTRATIONS--Continued

Figure

Quantity of Cobalt in Applied Effluent and in Filtrate

Samples During Lysimter Treatments

Quantity of Strontium in Applied Effluent and in

Filtrate Samples During Lysimeter Treatments

Page

82

83

ABSTRACT

Soil and grass filtration of a domestic sewage effluent for trace element removal was investigated by applying oxidation-pondtreated waste water to twelve 2.44-meter-long, 30.5-cm-diameter, lysimeter columns and to a half-acre plot planted to common bermudagrass.

Bermudagrass was also planted on the soil columns to aid in the removal of accumulated metals, thus prolonging the filtering capacity of the soil system.

Four irrigation patterns, continuous flooding, alternate day flooding, one day wet-three days dry cycles and two days wet-five days dry cycles, were employed during the eleven week lysimeter test.

Water samples were extracted from the soil columns at eight sampling depths to determine the soil depth at which the various trace metals were removed from the filtrate by the processes of adsorption, absorption and biological assimilation.

Atomic absorption spectrophotometric techniques were used to determine the concentrations of iron, manganese, hexavalent chromium, nickel, copper, zinc, lead, cadmium, cobalt and strontium in the applied effluent, filtrate, soil and grass clippings.

Filtrate samples for trace metal analysis were also extracted at shallow depths and from two groundwater zones at

9.1 and 15.2 meters depth under the grass plot.

In the lysimeter study, iron, manganese, nickel, copper, zinc, lead and cadmium were removed from the filtrate at, or near, the soil surface.

Some copper, zinc and cobalt were found in the deeper ix

filtrate samples during periods of high infiltration rates and in the presence of anaerobic conditions.

Smaller amounts of manganese, x nickel and lead were also found at the deeper sampling points.

Strontium was not removed from the water percolating through the calcareous soil employed in this study.

Chromium and cobalt were not present in measurable quantities in the applied effluent.

Iron and manganese were removed from the soil system in the grass clippings in the greatest amounts.

Traces of copper, zinc and cadmium were also detected in the grass clippings.

Larger quantities of iron, manganese and copper were detected in the filtrate at the grass plot than at comparable depths of the lysimeter columns, probably due to effluent applications at the grass plot over a period of two years.

Lower quantities of nickel and lead in the filtrate at the grass plot were attributed to leaching by effluent during grass filtration tests, prior to the trace metal experiments.

The concentrations of trace metals in the filtrate at the bottom of the lysimeter columns and in the groundwater zones at the grass plot were irsignificant by United States Public Health Service drinking water standards (1962) and would meet most other water supply requirements.

The total nitrogen and nitrate contents were successfully reduced when a combination of aerobic and anaerobic environments were present in the soil system.

Total nitrogen and nitrate levels were not reduced by grass filtration through 304.8 meters of dense grass.

xi

Fewer total coliform organisms were detected in the aerobic lysimeter columns than in the continuously flooded columns.

The majority of coliform organisms detected in the filtrate at the 61 cm depth were believed to be of non-fecal origin.

The chemical oxygen demand of the applied effluent was reduced to as low as 105 mg/i by grass filtration.

A substantial portion of the remaining COD was attributed to high algal concentrations.

The most effective treatment employed in this study was the one day wet-three days dry irrigation cycles.

This treatment provided the necessary aerobic environment for trace metal removal from the effluent, nitrification of reduced nitrogen compounds and coliform organism elimination.

The flooding period provided the anaerobic conditions required for denitrification losses of nitrate.

INTRODUCTION

Sewage effluent is a prime source of reusable water, provided economical and adequate quality treatments are available for reclamation.

Two of the several tertiary treatment methods for waste water reclamation are soil and grass filtration.

Purification processes in soil filtration systems occur primarily at the soil surface or in the biologically and chemically active surface layer of soil with lesser contributions from subsurface horizons.

Mechanical filtration of coarse organic materials occurs at the soil surface where the accumulated debris may be rapidly decomposed, mainly by aerobic organisms.

The dissolved solids in the percolate are removed near the soil surface by the processes of adsorption, absorption and biological assimilation.

It is believed that cations and some anions in the water are retained on the soil exchange sites where they may subsequently become available to plants and microorganisms.

Plant assimilation of ions regenerates the soil exchange capacity making the soil filter an efficient and durable renovating system.

In general, grass filtration is not as effective as soil filtration but it can be employed as a useful tertiary treatment for certain kinds of waste waters.

In grass filters, reclamation is accomplished by mechanical filtration of suspended solids through the dense gras.s and through utilization of the organic and inorganic substances by the biological films formed on the grass.

1

From an economic viewpoint, soil and grass filters should require minimum land areas, withstand maximum hydraulic loadings over long periods of time consistent with the desired quality improvement, require a minimum of maintenance and operational work, and be located near the waste disposal facilities.

2

Most grass and soil filtration studies for waste water reclamation indicate that reductions in nitrogen compounds, phosphates, calcium, magnesium, potassium, microorganisms, surfactants and other constituents can be accomplished by cation and anion exchange, microbial decomposition and plant assimilation but only limited studies concern the effects of reclamation on the trace elements.

These same processes should be effective in removing some, or all, of the trace metals from waste waters passing through the soil and grass filters.

The effect that accumulation of these ions in the soil could have on the filtering system is not fully understood but salts of some heavy metals are known to inhibit microbial activity and plant growth and, therefore, must be considered as vital factors in an efficient soil filter.

The objectives of the study reported herein were to determine

(1) the effects of soil and grass filtration on reducing the concentrations of trace elements in sewage effluent, (2) the effects of soil and grass filtration on other water quality criteria, (3) the floodingdrying cycle or cycles that provide the necessary treatment and (4) the quantity of domestic sewage that could be reclaimed per unit land area.

In order to achieve these objectives intensively instrumented lysimeter columns and a one-half acre grass filtration plot were established.

3

Various hydraulic loadings of sewage effluent were applied to these facilities and the following trace elements were monitored: iron, manganese, chromium (hexavalent), nickel, copper, zinc, lead, cadmium, cobalt and strontium.

These metals are contaminants in domestic, industrial and/or agricultural water supplies.

Water samples were also analyzed for nitrates, nitrites, ainmonium, organic nitrogen, chemical oxygen demand and coliform organisms.

LITERATURE REVIEW

Land filtration systems, for reclamation of municipal and industrial waste waters, have been used with varying levels of success during the past few decades.

In some instances the ultimate goal was to renovate the effluent so that it could be discharged directly into streams, rivers, lakes or oceans without degrading the quality of the natural waters.

Studies of this type have been reported by Searle

(1949), Foster et al. (1955), Porges and Hopkins (1955) and Parizek

etal. (1967).

Secondary benefits of many of these studies include the growing of crops, pasture grasses, forest trees and the recharge of groundwater aquifers.

The primary objective of work reported by Day and Tucker (1959) was to reclaim and utilize sewage effluent from the City of Tucson

(Arizona) Treatment Plant for the irrigation of several nonedible crop species with groundwater recharge occurring as an additional benefit.

Treated effluents have been used for recreational facilities such as the Santee recreation project, Santee, California (Merrell etal., 1965).

Laverty et al.

(1961) used a soil filtration system as a tertiary treatment method for improving the quality of Hyperion Treatment

Plant effluent so that it could be injected into the groundwater aquifers underlying the Los Angeles basin.

Where groundwater reservoirs are directly recharged through wells or indirectly by percolation through the overlying soil and rock strata, the reclaimed water must

4

be of the best possible quality to insure high quality waters for any future use.

5

General Waste Water Reclamation

Characteristics

To date, most waste water reclamation studies concern the reduction of nitrates, phosphates, alkyl benzene sulfonate

(ABS), chemical and biochemical oxygen demand (COD and BOD) and microorganisms from sewage effluents by soil or grass filtration with only limited references to the disposition of heavy metal ions.

By maintaining an aerobic soil environment, Parizek et al.

(1967) achieved excellent reductions for all elements tested with the exception of chlorides.

Only small nitrate losses from the filtrate were observed in the surface 30 cm soil layer but groundwater samples from beneath the irrigated plots contained only two to three mg/l of nitrate-N after 15 months of effluent irrigation.

They applied treated municipal waste water from the State College (Pennsylvania)

Treatment Plant to crop and forest lands by sprinkler irrigation at the rate of 5.1 cm per week.

After 18 months of waste disposal, during which time 289.6 cm of effluent were applied to the treatment area, the percolate samples extracted from the soil at a depth of 15 cm in a red pine (Pinus resinosa) forest contained 0.15 ing/l P, 12.1 mg/l NO3-N,

1.3 mg/l organic-N, 10.8 mg/i K and 77.3 mg/l Cl compared to 7.8 mg/l P,

13.6 mg/l NO3-N, 3.6 mg/l organic-N, 15.7 mg/i K and 39.6 mg/l Cl in the applied effluent.

Similar results were obtained for the chloride ion by Wheatland and Borne (1961) using Trent River water in England.

They also found little or no reduction in calcium, magnesium and

6 sulfate due to soil filtration.

During the initial phase of their flooding cycles the nitrate content varied from eight to ten mg/i, but decreased to six to seven mg/i as the environment became anaerobic.

The ammonia cQntent of the applied water was initially reduced, due to the oxidation of nitrogen, hut as the oxygen in the soil was depleted the ammonia content of the percolate approached that in the effluent.

Stone and Garber (1951) attributed nitrogen and phosphate removal from percolating water to "biological and chemical fixation." In subsequent work in the Los Angeles area, Stone (1953) found no pollution indicators in well water samples and concluded that land disposal of waste water can be effectively employed for the removal of various cations and anions provided intermittent applications are used to maintain an aerobic system.

Alkyl benzene sulfonate (ABS) is probably the most thoroughly tested organic compound in waste water reclamation studies and has been found to be one of the most difficult to eliminate from water percolating through the soil.

In an extensive study, Robeck etal. (1964) have shown that the most effective removal was accomplished in sand materials containing high amounts of carbon.

Applications of one gallon per day per square foot (gpd/sq ft) resulted in the removal of more than 90 percent of the ABS in the septic tank effluent, during the first 30 days of treatment, but when the dosage was increased to five gpd/sq ft only

70 percent of the ABS was removed during the same period.

Parizek et al.

(1967) reported a 97.7 percent reduction in the ABS content of their activated sludge and trickling filter effluent at 15 cm under the red pine forest during the first year but only a 93.1 percent decrease

during the second year even though the average ABS concentration in the applied effluent was reduced from 3.2 mg/i to 1.8 mg/i during the second year.

Robeck et al. (1964) concluded from their study that hydraulic loading is a definite limitation on the use of soil systems for waste water reclamation but that soil filtering systems can be used to trap and decompose ABS, as well as the usual COD components,

7 provided the surface soil layer has a low permeability and a high adsorptive capacity to hold suspended and dissolved organic materials.

Biochemical and chemical oxygen demand values are frequently used as a measure of the renovation capacity of grass filters and, in some cases, as a test for the oxygen requirements of percolate samples.

Biochemical oxygen demand is a measure of the amount of oxygen required by organisms in the biological decomposition of organic materials while chemical oxygen demand is the oxygen requirement for the chemical oxidation of all chemically oxidizable material.

Merreli et al.

(1965) accomplished BOD and COD reductions of 99 and 75 percent, respectively, by use of soil filters for reclaiming treated waste waters from the City of Santee, California.

Somewhat lower values are generally obtained for grass filters although

Truesdale et al.

(1964) report BOD reductions by grass filtration in

England equal to those obtained by slow sand filters.

A BOD reduction of sedimented sewage from 322 mg/i to less than 20 mg/i under hydraulic loadings of 17,300 gallons per acre per day was reported by Searle

(1949) for studies at Melbourne, Australia.

He attributed the reduction of organic pollution to oxygen dissolved at the water surface and to an active biological film built up on the herbage.

8

Stone (1953), Merrell etal. (1965),

and Bocko (1965) point out that Escherichia coil reductions exceeding 99 percent of the number in the applied effluent can be obtained by filtering the effluent through 120 to 210 cm of fine sand but the reduction is considerably less if the filter becomes anaerobic.

Bocko (1965) also listed "total bacteria" reductions of 85.0 to 99.6 percent under field conditions and 99.0 to 99.9 percent in lysimeters.

Complete virus removal from

Santee effluent was accomplished by soil filtration, as reported by

Merrell etal. (1965), but the majority of virus

was eliminated from the effluent in the primary and secondary treatments and the chlorination process prior to being applied to percolation beds.

Trace Element Filtration

Little work has been done in determining the removal of trace metals from water percolating through soil and grass filters.

Using river water, Wheatland and Borne (1961) reported considerable reductions in the concentrations of copper, nickel, total chromium, manganese, zinc and lead after percolation through the soil.

There was also a small decrease in the iron content.

Maximum reductions occurred in September and minimum reductions in January.

Ng and Bloomfield (1961 and 1962) studied trace metal mobilization and extractability in soil media and found lead, copper, nickel, zinc, cobalt and manganese to be mobilized by continuous flooding in the presence of fermenting plant material.

Reoxidation of the soil caused immobilization of all these metals except copper and nickel by the process of coprecipitation with ferric oxide.

Thomas et al. (1966)

reported that iron precipitated as ferric oxide coatings on soil peds at a logarithmic rate under aerobic conditions but was reduced to the ferrous form under continuous inundation.

Small amounts of chromium were also mobilized by flooding according to Ng and Bloomfield

(1962) and about half the chromium remained extractable after reaeration.

9

Ion exchange was presumed to be the primary process for removal of copper, chromium, zinc, lead, nickel, iron and manganese in work reported by Wheatland and Borne (1961).

Mitchell (1964) also reports that many trace metals are absorbed or complexed by the organic fraction, especially aluminum and ferric iron.

He reported copper, cobalt, and chromium chelates were found in the "fulvic acid fraction" of organic matter.

Plant and Organism Response to Trace Metals

The response of plant growth to macroelement uptake from sewage effluent has been well documented (Searle, 1949; Day etal., 1962;

Parizek et al., 1967) but information concerning plant response to the various trace elements is incomplete.

Substitution of elements has been reported by Thorne and Wiebe (1957) who found manganese absorption to be reduced by increased uptake of iron and other heavy metals and by

Hill et al. (1953) who reported cobalt uptake to be reduced by increased iron and manganese assimilation.

Beeson et al.

(1947) analyzed several grass species for trace metal content, including coimnon bermudagrass (Cynodon dactylon) grown in

Arizona.

They reported cobalt, manganese and copper contents of 0.07 ±

0.01 pg/g, 176.7 ± 9.63 pg/g and 5.3 ± 0.36 pg/g, respectively.

Riley

10

(1967) and his co-workers analyzed common bermudagrass samples obtained from one farm near Scottsdale, Arizona, by atomic absorption techniques.

The cobalt content of their samples varied from less than oqe to 2.3 pg/g, copper from 18 to 33 pg/g, iron from 585 to 1025 ig/g, manganese from 85 to 105 g/g and zinc from 42 to 57 pg/g.

The variability of nutrient content in plants is influenced by soil composition, pH, moisture status, fertility level and the trace element content of the plant (Mitchell, 1964).

In addition to trace element removal from the soil by deep percolation and incorporation into plant tissue, there is also some loss due to uptake by soil microorganisms.

However, this loss may only be temporary, the metals again being released upon lysing of the organisms.

Starkey (1955) includes iron, copper, manganese, zinc and cobalt among the trace elements used by the soil microorganisms.

The uptake of trace metals by plants may also be temporary unless the plants are harvested and removed from the area.

Techniques for Trace Metal Detection

Several methods have been used to measure trace element content of water, soil and plant material but, to date, the most accurate and most rapid method is atomic absorption spectrophotometry.

This method is based on the principle that atoms of metallic elements absorb light of the same wavelength they emit when excited (Platte and Marcy,

1965).

When radiation from a given element passes through a flame containing atoms pf that element, the radiation

intensity

is decreased in proportion to the amount of the element in the flame.

The radiation

11 source, a cathode made from the element to be determined, emits a radiation beam into which the metal atoms are placed by aspirating the samples in a gas flame.

The characteristic wavelength of the unexcited atoms from the cathode is isolated by a monochrometer and the unabsorbed radiation is measured by a photosensitive device.

This analytical technique was developed by Alan Walsh in 1955.

Price (1967) states that gas mixtures such as air-propane and nitrous oxide-acetylene have considerable advantages in a number of cases but air-acetylene is the most useful because temperature and reducing conditions can be varied within fairly wide limits.

Platte and Marcy (1965) state that atomic absorption is highly accurate because it is specific for a given metal, even in the presence of other metals.

Wet chemical methods and flame photometry fail in the presence of interfering ions.

The specificity of atomic absorption spectrophotometry for individual ions was shown by analyzing solutions containing one mg/i iron, copper, nickel, zinc, manganese, chromium, calcium or magnesium in the presence of 1000 mg/i of sulfate, chloride, phosphate, nitrate, nitrite, bicarbonate, silica, EDTA, iron, nickel, zinc, manganese, chromium, boron, lead, calcium, magnesium or sodium ions.

Results in the range of 0.97 to 1.02 mg/i were reported (ibid).

The detection limits of atomic absorption spectrophotometry for the various ietals are: zinc, 0.004 mg/i; copper, 0.008 mg/i; manganese, 0.012 mg/i; chromium, 0.015 mg/i; iron, 0.024 mg/i; and nickel, 0.032 mg/i (ibid).

Ulrich and Shifrin (1965) suggest that refinement of sample preparation techniques will aliow even lower detection limits.

Price (1967) indicates many elements found in

industrial wastes, including cadmium, chromium, copper, iron, manganese, lead, nickel, silver and zinc, are among those which show the greatest sensitivity in atomic absorption.

Platte and Marcy (1965) found excellent agreement between atomic absorption and colorimetric procedures for copper concentrations abQve 0.10 mg/i and iron concentrations above 0.025 mg/i.

The recovery of zinc and nickel was also excellent by atomic absorption but not by standard gravimetric methods.

12

MATERIALS AND METHODS

The effectiveness of a soil and grass filter for quality improvement of waste water was determined by percolating sewage effluent through the soil in a lysimeter and by flowing sewage effluent through a dense common bermudagrass cover and allowing the water to infiltrate into the soil.

Common bermudagrass was chosen because of a demonstrated tolerance to long periods of flooding (Wilson,

1967), ability to grow throughout most of the year in the southwest, the minimum amount of maintenance and its value as a forage crop.

The concentrations of the various metals in the sewage effluent applied to the filtering systems, and the subsequent disposition of these metals in he system, were determined by analyzing samples of the effluent, percolate, soil and vegetation.

These measurements allowed the determination of the quantities of trace metals removed from the percolating water by soil sorption and plant assimilation and the quantities that passed through the system in combined forms.

Auxiliary studies were conducted on the sewage effluent and filtered water to measure the efficiency of the system in terms of other water quality criteria.

Materials

Sewage Effluent Supply

The waste water used in these experiments consisted of effluent from two oxidation ponds, operated in series by the Sanitary District

13

No. 1 of Pima County.

These ponds are used to treat the sewage from the outlying residential areas north of Tucson, Arizona, before discharging the effluent into

the

Santa Cruz River.

The total storage

14 capacity of the ponds and an average inflow of 0.3 million gallons per day allow for a detention time of 30 to 40 days.

Treatment in the oxidation ponds is effective because of the favorable light intensities and warm yearlong temperatures which are necessary to establish an algal bloom.

The algae provide much of the oxygen required in the digestion processes.

The COD of raw sewage, approximately 400 mg/i, is reduced to 150 to 300 mg/i depending on the time of year.

Much of the remaining chemical oxygen demand can be attributed to the large quantities of algae in the outflow.

Grass Plot Facility

Three parallel grassed plots, each approximately 6.1 x 304.8

meters with a uniform slope of 0.3 percent, were constructed contiguous to the oxidation ponds, ten miles north of Tucson.

The sewage effluent overflow from the ponds was emptied into a transite-lined head ditch along the end of the plots from which the water was diverted, through flumes, onto the plots.

An unlined tail ditch carried the runoff from the strips to the river.

All instrumentation at the site was located on the central of the three plots, the outer plots serving as buffer zones for percolating waters.

Instrumentation included (1) 7.62 cm Parshall flumes at the head and tail ditches with water level recorders for measuring flow on and off the plots, (2) 4.45 cm (ID) seamless steel access tubes at

15

61.0 and 243.8 meters from the head ditch, installed to a depth of

30.5 meters, for use with a neutron soil moisture logger, (3) three sets of soil moisture sampling tubes 76.2, 152.4 and 228.6 meters from the head ditch at depths of 15.2, 30.5, 45.7 and 61.0 centimeters and

(4) 10.2 cm (ID) PVC plastic-lined sampling wells at 9.1- and 15.2meter-depths with l.5-meter-long well screens.

The two plastic-lined wells were installed after studying logs obtained at the two access tubes during preliminary investigations

(Wilson and Lehman, 1967).

All records indicated an upper zone of saturation between 7.9 and 9.4 meters and a groundwater table that fluctuated between 12.8 and 14.6 meters, depending on the volume and time of recharge at the site and on the amount of pumping at a nearby irrigation well.

The soil and geologic properties at the grass plot site were studied in detail by exposing a profile to 2.44 meters and conducting textural analysis and bulk density determinations for each 5.1 cm increment.

Below this depth, properties were determined from the drill cuttings obtained during installation of the wells.

Textural analysis was accomplished by the wet sieve method (Day, 1965) for all particles with an effective diameter greater than 0.05 mm and by the hydrometer method for clay particles (ibid).

Distinctly stratified alluvial deposits underlie the entire grass plot area.

The surface 25 cm of soil is primarily composed of very fine sand and silt-sized particles with less than 8.5 percent clay (Table 1).

At the 25 cm depth there is an abrupt change to medium and fine sand with even smaller amounts of clay material and

16

Table 1.

Soil Textural Analysis of Distinct Horizons in Situ at Grass Plot

Site for Soil

Depth Gravel

Very Coarse

Medium

Coarse

Sand Sand

Sand

Fine

Sand

Very

Fine

Sand

Silt Clay cm

>2

mm

1-2

mm

.5-1

mm

.25-.5

mm

.l-.25

mm

percent

.05-.1

mm

.002-.05

mm

<.002

mm

0-25

0.8

1.2

3.6

7.5

16.3

26.5

35.7

8.4

25-61

0.2

1.8

9.4

23.8

27.9

18.0

11.2

4.7

61-76

0.1

0.5

1.4

3.1

6.7

15.2

60.1

12.9

76-127 0.3

1.5

1.0

3.1

25.8

30.8

30.8

6.7

127-137

2.7

8.0

16.3

17.2

11.7

16.6

20.8

4.4

137-152 10.7

21.8

25.2

18.4

10.8

3.3

6.6

3.8

152-163

26.4

163-168

7.8

17.2

10.5

9.3

14.3

9.7

20.2

6.9

13.7

4.7

6.3

18.9

22.2

5.7

6.2

168-203

24.8

28,5 23.8

12.9

3.9

0.6

1.0

4.5

203-218 6.9

21.6

35.1

23.3

4.2

0.3

3.2

5.4

218-224

224-239

4.7

1.1

239-244

14.9

13.8

19.8

1.5

6.0

22.1

21.2

32.6

50.5

31,1

19.6

29.8

10.4

2.5

3.4

0.8

0.8

1.5

*

6.2

6.2

2.5

*The silt content was a negative value dueto error caused by using wet sieve analysis for the coarser fractions and the hydrometer method for the clay particle content,

17 only about one-third the silt content of the overlying stratum.

The clay and silt contents reach a maximum, in the 2.44 meter profile, between 61 and 76 cm with values of 12.9 percent and 60.1

percent, respectively.

Very little material larger than fine sand exists in this layer.

Fine sand, very fine sand and silt predominate in the 76 to 127 cm stratum, the deepest uniform layer in the surface 2.44

meters of the profile.

Nearly equal amounts of coarse, medium, fine and very fine sand, with a somewhat larger silt content, characterize the 127 to 137 cm layer.

Very coarse sand and some fine gravel replace the silt-sized particles between 137 and 152 centimeters.

Two more thin beds, 152 to 163 cm and 163 to 168 cm, contain large amounts of fine gravel with coarse sand material and medium sands with little gravel or coarse sand, respectively.

Coarse sands and fine gravels predominate from 168 to 203 cm with only minor amounts of the finer materials.

Medium to very coarse sand occur between 203 and 218 cm.

A transition layer between 218 and 224 cm, with slightly less coarse sand but more medium and fine sands, overlies a 15 cm stratum of medium and fine sands from 224 to 239 cm.

A gravel, coarse and medium sand layer extends from 239 cm to the base of the exposed profile.

Alternating beds of coarse and fine material were found in the drill cuttings to a depth of 15.2 meters.

A layer of fine material was located between 9.4 and 10.7 meters.

Restricted drainage through this layer appeared to be responsible or the upper saturated zone between

7.9 and 9.4 meters under the plot.

18

Common bermudagrass was planted on the strips two years prior to the initiation of the trace metal experiments and had become well established.

An excellent filter was formed by allowing the dense grass to grow to a height of 35 centimeters prior to testing.

This procedure reduced the possibility of channeling and inundation during flooding cycles.

Lysimeter Construction

A more detailed study of soil filtration was accomplished by constructing soil columns at the Water Resources Research Center Field

Laboratory, about eight miles south of the grass plots site.

All materials used in the construction of the lysimeters and instruments for obtaining water samples were checked for adsorptive and desorptive effects on trace elements.

The plastics, polyethylene and pyrex glass used were found to have little or no adsorptive or desorptive properties.

The lysimeter consisted of 12 soil columns arranged around the periphery of a 1.5-meter-diameter steel culvert pipe set vertically below the soil surface (Figures 1 and 2).

ABS plastic tubing, with a diameter of 30.5 cm (ID) and a length of 259.1 cm, was used to form the soil columns.

The 30.5 cm diameter columns provided sufficient area for the growth of grass on the surface and reduced the soil-column wall interface effect.

The common bermudagrass planted on the columns was irrigated with deionized water prior to the sewage effluent filtration tests.

The open-bottomed columns rested on soil material similar to that contained within the columns.

This procedure simulated field

SEWAGE EFFLUENT

[SUPPLY RESERVOIR

SOIL

COLUMN

PIEZOMETER

TUBE

(THERMOMETER

TUBE

VACUUM MANIFOLD

FILTRATE

COLLECTION

BOTTLE

F LOAT

VALVE

WELL

FILTRATE

SAMPLING

TUBE

FIGURE

1.

Lysimeter

Cohstruction

Effluent

Distribution

Showing Arrangement

System and or

Instrumentation.

Soil

Columns,

19

TOP OF

CULVERT

I

MANOMETER

PANEL

THERMOMETER

TUBES

VACUUM

MANIFOLD

\

F I LTR ATE

SAMPLING

TUBES

FILTRATE

COLLECTION

BOTTLE

- TENSIOMETER

VACUUM

TANK

CONCRETE PAD

BOTTOM OF

CULVERT

\

00 op

0 0

FACE VIEW

TYPICAL SECTION

OF SOIL COLUMNS

FIGURE 2.

Face

View of Inside of Lysimeter Collection Chamber and

Typical Section of

Soil Columns.

SOIL

LEVEL

PIE ZOMETER

TUBE

SOIL

COLUMN

20

21 conditions were suctions result from the continuity of moisture films down to the water table.

Use of the same textured material below the soil columns also assured the continuous drainage of the columns without an interface effect due to a soil textural change and eliminated the problem of entrapped air caused by unnatural drainage.

The soil material used in the columns was obtained from the surface 2.44 meters adjacent to the grass plots.

Soil material, collected in 30.5 cm depth increments, was screened to remove rock material then thoroughly mixed.

To reduce settling in the columns with time, and simulate field conditions, packing of each 30.5

cm layer was based on the previously obtained bulk density values of the soil in situ (Table 2).

The clay content of all layers was between 4.0 and 6.5 percent with the exception of the 61 to 91 cm layer which had a clay content of 9.6 percent.

The silt content was greatest in the upper 122 cm with a maximum 48.3 percent between 61 and 91 cm.

Below 122 cm the silt and clay fractions were only a small part of the soil composition.

Approximately 20 percent of the material between 152 and 213 cm was fine gravel.

The total cation exchange capacity of the soil material in the lysimeter columns, determined by the sodium acetate method at pH 8.2, was 7.5, 7.1, 7.1 and 7.1 meq/lOO grams of dry soil for the 0.0 to 7.6,

7.6 to 15.2, 15.2 to 22.9 and 22.9 to 30.5 cm depths, respectively.

The material between 30.5 and 61.0 cm had a cation exchange capacity of

9.7 meq/100 grams of dry soil.

22

Depth

Table 2.

Textural Composition and Bulk Density of Soil

Material Packed in Soil Columns

Sand and

Fine Gravel

Silt Clay Bulk

Density cm

0-30

30-61

61-91

91-122

122-152

152-183

183-213

213-244

>.05

mm

61.6

83.0

42.1

63.8

79.2

83.7

96.3

92.1

.002-.05

mm percent

31.9

12.4

48.3

31.2

16.8

12.1

*

2.5

<.002

mm

6.5

4.6

9.6

5.0

4.0

4.2

5.8

5.4

g/cc

1.20

1.30

1.27

1,24

1.28

1.35

1.38

1.29

*The si't content was a negative value due to error caused by using wet sieve analysis for the coarser fractions and the hydrometer method for the clay particle content.

23

Five centimeters of each plastic tube protruded above the soil surface to provide a reservoir for the applied effluent.

The reservoir was connected to a shallow 10,2 cm diameter PVC plastic stilling well by a short tube.

Specially designed plexiglass and plastic floatactivated valves were connected in series of six to the plastic-lined

208 liter drums that served as the effluent supply tanks.

The float valves maintained a constant water level on the soil columns.

Stirring mechanisms were installed on the supply drums.

All recording and filtrate sampling devices used in the soil columns terminated within the central collection chamber.

Instrumentation included three piezometer tubes, a tensiometer, eight filtrate sampling tubes and ten thermometer access tubes in each column.

The piezometers consisted of 6.35 mm plastic tubing, sealed at one end and drilled with fine holes, connected to glass manometer tubes inside the chamber,

The tubes were located at 0.0, 7.6, and 15,2 cm depths in the soil columns.

The tensiometer was installed 2.44 meters below the soil surface and consisted of a porous ceramic cup attached to the end of a rigid plastic tube.

A fine nylon tube connected the cup to a mercury manometer inside the central chamber.

Water sampling devices were located at depths of 7.6, 15.2,

22.9, 30.5, 61.0, 121,9, 182.9 and 243.8 cm below the soil surface.

Porous ceramic cups located at the center of the columns and mounted on rigid plastic tubes, similar to the tensiometers, were used for extracting water samples.

The method employed for obtaining water samples was described by Reeve and Doering (1965).

All eight sampling

tubes within each column were connected to a single vacuum cylinder

24 by a vertical manifold inside the central chamber.

Methods

Effluent Application Rates

The supply of sewage effluent for all lysimeter columns was collected at the oxidation pond overflow pipe each morning.

The daily sample allowed for a more precise description of the effluent being applied then would have been possible using a continuous supply.

The stirring mechanisms in the supply reservoirs provided the necessary agitation to maintain a homogeneous sample throughout the day.

The supply tanks were drained and refilled with fresh effluent at noon each day.

The sewage effluent was applied to the lysimeter columns by maintaining a three to four centimeter head of water on the columns during irrigation.

Four different irrigation patterns were used during the study.

Initially, half of the twelve columns were continuously flooded and the remaining six were flooded on alternate days.

After six weeks, three of the columns in the continuously flooded group were placed on a two days wet-five days dry cycle while continuous inundation was maintained on three columns for the final five weeks.

Alternate dry flooding was continued on three of the six columns receiving that treatment while three columns were placed on a one day wet-three days dry cycle.

25

Sampling

Samplig procedures for trace elements.

Only a limited number of samples could be analyzed for trace metal content due to the complexity of sample preparation for trace metal determinations and because changes in metal concentrations, by adsorption and desorption in the containers, limited the storage time.

With these restrictions in mind, the following procedures were employed for sampling the applied effluent, filtrate, soil and vegetation.

Samples collected daily from the effluent supply reservoirs were taken to be representative of the waste water applied to the lysimeter columns.

The samples were made into three and four day composite samples for trace metal analysis by combining equal volumes of the daily samples.,

Initially the filtrate samples were collected at all eight sampling depths but after filtering trends were established the number of sampling points was reduced.

During the final five weeks, samples were obtained at depths of 7.6, 15.2, 22.9, 30.5 and 243.8 cm.

Semiweek filtrate samples, from each sampled depth for all columns having the same flooding cycle, were made into composite samples for trace metal analysis.

The bermudagrass was cut prior to irrigation and analyzed for the various trace metals.

During the study, the grass was cut biweekly at five centimeters above the soil surface.

The clippings were combined into composites for each of the treatments.

Prior to irrigation, soil sdmples were obtained from six randomly selected columns at depths of 0.0 to 7.6, 7.6 to 15.2, 15.2 to

26

22.9, 22.9 to 30.5 and 30.5 to 61.0

cm.

Samples were also collected, after the first six weeks of irrigation, from all treatment sets of columns except the continuously flooded columns.

At the termination of the study, soil samples were collected from the top 61.0 cm of each of the four treatments.

Composite samples were made for each soil layer, for every treatment.

Filtrate samples from the grass plot site were collected at depths of 15.2, 30.5, 45.7 and 61.0 cm at three points along the plot.

Samples collected at random during the two and five week tests were analyzed individually for trace elements.

Random water samples were also obtained from the 9.1 and 15.2 meter wells.

The quality of water samples extracted from the 9.1 meter well was indicative of treatment received in the soil and upper lithologic strata prior to mixing with groundwater while samples obtained from the 15.2 meter well had characteristics derived from mixing with native groundwaters.

Sampling procedures for nitrogen compounds, coliforms and COD.

Nitrate determinations were conducted on water samples extracted from the lysimeter columns at the 61.0 and 243.8 cm depths.

The samples, collected at random intervals during the test, were analyzed for each column rather than by treatment sets of columns.

During the tenth week of testing, ammonium, organic nitrogen and nitrite analyses were conducted in addition to nitrate.

Nitrate determinations were also made on random filtrate samples from the 15.2, 30.5, 45.7 and 61.0 cm depths at the grass plot.

Well samples, applied effluent and outflow from the grass plot were analyzed for nitrates, nitrites, ammonium and organic nitrogen.

27

Daily COD tests were conducted on the applied effluent throughout the three month study and on the outflow from the grass plot during the two and five week flooding experiments.

One inflow and one outflow sample per day were used for COt characterizations.

Coliform organism counts were made on filtrate samples from the 61.0 and 243.8 cm depths of the lysimeter columns and from the 15.2,

30.5, 45.7 and 61.0 cm depths at the grass plot.

The samples were collected at random times during the tests.

Chemical Analyses

Sample preparation for trace metal analysis.

The residue remaining after evaporating 100 ml aliquots of the three and four day composite effluent samples was digested using the nitric and perchloric acid method outlined in Standard Methods (American Public Health Association, 1965) for heavy metals.

The digested material was diluted to volume and filtered with .1 N HC1-washed Whatman No. 30 paper, the filtrate being used for atomic absorption spectrophotometric analysis.

Filtrate samples from the lysimeter columns and grass plot were filtered through acid washed Whatman No. 30 paper before trace metal determinations were conducted, but no further treatments were employed.

Soil samples were combined for each treatment and dried at

105°C for 24 hours before extraction with sodium acetate.

One hundred grams of dried soil, placed in a Buchrier funnel, was rinsed with two

50 ml aliquots of 60 percent methanol and the leachate collected in a vacuum flask.

Replacement of ions on the exchange site was done by

28 using three 50 ml washes of 1 N sodium acetate solution at pH 8.2.

The leachate was also collected in the flask.

Three final 50 ml 60 percent methanol washes were used to leach the replaced ions from the soil.

The combined leachate from the eight washes was transferred to a beaker and evaporated over low heat until dry.

The cooled residue was predigested with 10 ml of nitric acid before a final digestion with 10 ml of perchloric acid.

The remaining material after digestion was diluted to volume and filtered through acid washed Whatman No. 30 paper, the filtrate being used for trace metal determinations.

The total grass clippings from each treatment were combined and oven-dried at 66°C for 24 hours.

After the total dry weight was determined, a 0.400 gram sample was obtained for each treatment.

The grass samples were digested by the nitric-perchloric acid method described for the soil extract, followed by dilution and filtration.

Metal extraction techniques for atomic absorption analysis.

Special trace element extraction techniques were employed for preparing the liquid samples for atomic absorption spectrophotometric analysis.*

Extraction from the liquid samples was accomplished, for all ions except strontium, by adding an ammonium acetate-acetic acid solution, methylisobutyl ketone and sodium diethyldithiocarbamate to the liquid samples.

The sample was stirred before being diluted with deionized water and analyzed.

*The extraction techniques for analysis of the various trace elements contained in the liquid samples were developed by T. W.

McCreary and G.

R. Dutt of the Department of Agricultural Chemistry and Soils, University of Arizona, Tucson, Arizona.

29

Strontium was extracted by adding solutions containing three different concentrations of all trace metals included in the study, to three aliquots of sample.

Lanthanum oxide was added to each sample prior to analysis on the atomic absorption unit.

Chemical analyses for nitrogencornyounds and COD.

All nitrate samples were analyzed by a modification of the brucine method.

Twelve drops of brucine were added to five ml samples followed by ten ml of sulfuric acid,

Ten ml of deionized water was added exactly five minutes after the acid.

The flasks were allowed to cool for ten minutes before the solution was transferred to a test tube.

The transmittance was read at 410 mi.i and the nitrate contents determined from a standard curve developed from samples containing known concentrations of nitrate.

The Kjeldahl method (Bremner, 1965) was used to measure organic nitrogen and steam distillation methods (ibid) were used to determine nitrites, nitrates and ammonia nitrogen in the effluent, the runoff from the grass plot and water samples from the 9.1 and 15.2 meter wells.

Chemical oxygen demand determinations were made using the technique described in Standard Methods (1965).

Fifty ml samples were used in the COD tests for the applied effluent and the outflow from the grass filter.

Coliform Enumeration

Coliform organism counts were made by drawing water samples through a 0.45 micron Millipore filter with a suction flask.

The filter was placed on an absorbent pad saturated with Endo broth and

30 placed in a sterile petri dish.

Counts were made after 48 hours incubation at 35 ± 0.5°C.

Physical Measurements

Daily infiltration rates for all columns being irrigated were obtained by measuring the head losses in the stilling wells during a given period of time.

Measurements were made with a point gauge.

The head loss in millimeters for the stilling well and soil column combined was converted to loss in the column only by multiplying by the factor of 1.11 (total volume loss in column and stilling well divided by the area of the column).

Piezometer readings were obtained at the same time the infiltration rates were recorded.

Tensiometer measurements were recorded daily for all columns in order to establish drainage patterns within the columns.

Tensiometer response provided information concerning the initial rates of water percolation through the 2.44 meter columns.

Temperature measurements were taken periodically during the study at the various soil depths and at the soil surface.

Precipitation, temperature and evaporation data were obtained from the Water Resources Research Center weather station located within

100 meters of the lysimeter columns.

Data Analysis

Calculation of application and infiltration rates.

The volume of effluent applied to each lysimeter column was estimated by dividing the volume of effluent withdrawn from a supply reservoir by the percentage calculated for each column receiving water from that reservoir.

The percentage allotments were based on the measured daily infiltra-

31 tion rates of the individual soil columns,

The losses from the columns were estimated from evaporation pan data, which reflect changes due to evaporation and precipitation.

Daily evaporation pan losses, in centimeters, were converted to liters lost from each soil column by the factor of 0.247 liters per cm (area of soil column + area of stilling well - area of float = 247 sq cm x

1 cm depth = 0.247 1/cm).

Trace element calculations.

The quantities of trace elements applied in each treatment were calculated from the metal concentrations detected in the composite effluent samples and the estimated application rates.

The metal concentrations from the atomic absorption unit in mg/i, for the applicable composite sample, were multiplied by the volume of waste water applied to each treatment for every day of irrigation included in the composite sample period.

Since filtrate samples, and therefore, trace metal concentrations, were only obtained biweekly it was necessary to extrapolate between sampling days to obtain trace metal concentrations for days when filtrate samples were not collected.

The measured and extrapolated daily values were multiplied by the estimated infiltration rates to obtain the quantities of trace metals in the filtrate at the various sampling depths.

The concentrations in milligrams of trace metal per 100 grams of dry soil were converted to milligrams of the various metals in each soil layer sampled by multiplying by the volume of the layer and the bulk density and dividing by 100.

32

The results of the grass analysis were converted to milligrams of trace metal removed in the clippings by multiplying the mg per gram dry weight by the dry weight of the grass removed.

RESULTS AND DISCUSSION

The reclamation of sewage effluent by soil and grass filtration was studied by irrigating 12 lysimeter columns and a half-acre grassed plot.

Of primary concern was the removal of trace elements from the waste waters and the uptake of these metals by the grass growing on the surface.

Secondary studies included determining the changes in nitrogen concentrations, coliform organism counts and chemical oxygen demand.

Effluent Application Rates, Losses and Infiltration Rates

In the lysimeter study, 12 soil columns were arranged into four units of three columns each.

During experimentation the three columns in each unit were treated alike.

The treatments applied were

(1) continuous irrigation for 11 weeks (designated C),

(2) continuous irrigation for six weeks followed by two days wet-five days dry cycles for five weeks (designated C-2F5D), (3) alternate day irrigation for

11 weeks (designated 1F1D) and (4) alternate day irrigation for six weeks followed by one day wet-three days dry cycles for five weeks

(designated 1F1D-1F3D).

Since chemical analyses were conducted on water samples from each set of three columns as a unit, rather than on individual columns, the effluent application data is presented in a similar manner.

Effluent application rates for each of the four treatments presented in Figures 3, 4, 5 and 6 are the summation of the

33

300

200

1hhIL

0

1

2

3

4

5

6

WEEK

7

8

0

FIGURE 3.

Total Volume of

Effluent Applied Daily to Continuously

Flooded Columns.

34

300

200

'U

-J

0

U4

La-

0

Iz

LU

D U)

-J u_

Uu.I

u.j

-

-j

-

D

0

>

100

0

WEEK

FIGURE 4.

Total Volume of

Effluent Applied

Each

Jrrigation to

C -2F5D

Columns.

Day of

35

FIGURE 5.

Total

Volume of Effluent Applied

Irrigation

to

1F1D

Columns.

Each

Day of

36

300

-J

200

U, uJ

I-

U-

0

UJ

-J

0

>

100

0

1

2 3

4 5 6

WEEK

7

8

9

10

FIGURE

6.

Total Volume of Effluent

Applied

Each

Day of

Irrigation to

1HD - 1F3D

Columns.

11

37

38 volumes of effluent applied per day to the three columns receiving a given treatment.

The estimated losses, corrected for rainfall, are listed in

Table 3 along with the application rates and estimated infiltration rates.

The total evaporation loss for the three continuously flooded columns during the 11 week period was estimated as 160 liters, or

6.8 percent, of the applied water.

Losses for the three other treatments were less since free water was standing on the surface for only a fraction of the total test period.

The infiltration rate values approximate the reported application rate curves, the infiltration values being slightly less than the applicatioTi rates.

During the first six weeks of testing, the six continuously flooded columns received a total of 3928 liters of effluent of which an estimated 3747 liters infiltrated Into the soil.

The remaining 181 liters was lost by evaporation.

A total of 3218 liters was applied to the six lFlD columns during the same period with an estimated 3035 liters infiltrated into the soil and 183 liters evaporated.

The three 1F1D columns received 1084 liters during the final five weeks compared to 1218 liters and 1402 liters for the 2F5D and

1F3D columns, respectively.

Only 564 liters of effluent were applied to the continuously flooded columns during this period.

The application rates in each of the two initial treatments were high but decreased rapIdly during the first few days of irrigation.

For the continuously flooded columns application rates decreased rapidly in the first three days of testing, from 237 and 195 liters per

day to 57 liters per day, followed by lower and irregular rates for the remainder of the first two weeks.

The application rates doubled

40 at the end of the two week period and remained at this level for more than a week before beginning a gradual two-week decline to approximately

15 to 20 liters per day.

Although the rate dropped to ten liters per day, it reached a plateau around 15 liters per day for the final weeks of the test.

Conversion of the continuous flooding pattern, on one set of three columns, to a 2F5D cycle resulted in an increase in sewage application rate from approximately 40 liters per day to 117 liters per day on the first day of irrigation following the initial drying period and an ultimate rate of approximately 160 liters per day by the termination of the test.

Application rates on the two sets of columns that received alternate day flooding fluctuated between 50 and 80 liters per day after the first week of irrigation.

The application rate for the three columns maintained on alternate day flooding for the entire 11 week period remained, for the most part, at the 50 to 80 liters-per-day level until the final three weeks when the application rates fluctuated between 45 and 53 liters per day.

The second set of lFlD columns failed to maintain the 50 liters-per-day rate after three weeks of irrigation, partially due to a sparser grass cover than on the other set of 1F1D columns.

Application rates decreased to approximately 25 liters per day during the fifth week of treatment but recovered to approximately

125 liters per day as soon as flooding was reduced to every fourth day.

41

The greatest volume of effluent was applied when drying periods were used between inundations.

Maximum effluent application rates occurred on the columns employing the 1F3D cycle and a greater volume of effluent was applied in this treatment than in the 2F5D treatment, during the final five weeks of the study.

It must be remembered that when intake rates are high, the concentrations of contaminants introduced into the system are also high.

The total amount of any component in applied sewage effluent may be the limiting factor ultimately determining the flooding cycle to be used in soil filtration of sewage effluent for quality improvement.

Organic mats were observed on the soil surface of the continuously flooded and alternate-dry-flooded columns but not on the 2F5D or 1F3D columns.

The failure of organic materials to accumulate on the columns having drying periods of three and five days duration was possibly the result of rapid oxidation of any organic substances on the surface when exposed to air.

The accumulated material on the surface of the C and 1F1D columns could have been the primary factor influencing the infiltration losses observed for these treatments.

Laverty etal.

(1961) attributed losses in infiltration rate to an organic mat formed on the soil surface during continuous irrigation with a mixture of

Hyperion effluent and groundwater.

They reported that when the basins were dried, and the organic material exposed to the sun and air, the seven to ten cm thick mat disappeared within two days leaving a surface layer of clean sand.

42

Ph sical Environment of Lysimeter

Columns

The probable aeration and drainage conditions of the 12 lysimeter columns were estimated from the nitrogen transformations, tensiometer data and soil textural analysis.

Although the auxiliary nitrogen studies will be discussed in detail in a later section, general trends in nitrate and total nitrogen contents will be presented here as evidence for the aeration status of the soil columns.

A combination of an aerobic and an anaerobic environment is necessary for the elimination of nitrogen from the soil system

(Alexander, 1961).

The nitrification of reduced nitrogen compounds requires the presence of oxygen but even small amounts of oxygen are sufficient to cause some ammonium oxidation.

However, the reaction ceases in the total absence of oxygen (ibid).

Decreases in nitrate content are the result of denitrification, immobilization, leaching losses and reduction to ammonia (Broadbent,

1951 and Broadbent and Stojanovic, 1952).

Denitrification to nitrogen gas and nitrous oxide is the dominant process for nitrate loss in high pH soils when the moisture content is greater than 60 percent of the water holding capacity of the soil and organic materials are present in the soil (Bremner and Shaw, 1958).

The organic material serves as an energy source for the microorganisms and as a hydrogen donor in the denitrification process (ibid).

Complete anaerobisis is not necessary for denitrification to occur.

The principal methods for reaeration of the soil are (1) dissolved oxygen carried into the soil by percolating water, (2) the movement of air ahead of the wetting front and (3) the diffusion of

43 air into the soil media when the soil surface is not inundated

(McMichael and McKee, 1966).

Under continuous irrigation, only small changes in the tensiometer recordings were noted throughout the study.

The soil mpisture tension at 243.8 cm increased from 33.8 to 35.0

cm of water between the second and seventh days of the study suggesting some drainage from the lower part of the soil columns during this period.

No further changes were observed during the final ten weeks.

Increasing amounts of nitrate in the filtrate with depth also suggest the possibility of some reaeration occurring in the lower part of the soil columns.

The filtrate samples extracted at the 61.0 cm depth had a slight increase in nitrate content over the amount of nitrate in the applied effluent, during the first day of the test.

This increase was attributed to the oxygen and nitrate in the soil system prior to irrigation.

The nitrate content of the filtrate decreased on the second day of irrigation as the oxygen was depleted.

Increases in the nitrate content at this sampling point during the course of the study was apparently the result of partial reaeration of the soil at, or above, this depth.

Greater nitrate contents in the filtrate at

243.8 cm indicates the presence of a more aerobic environment between

61.0 and 243.8 cm than that found in the surface 61.0 cm.

The sand and fine gravel content, more than 79.2 percent, could have allowed more drainage from the soil columns below 122 cm depth than from the finer materials in the upper soil layers.

Apparently a similar phenomenon occurred within the top 61 cm layer of soil as the study progressed.

Partial reaeration of the coarser material ia the 30.5

to 61.0 cm layer allowed some nitrification of the reduced nitrogen compounds to occur, even though the surface soil layer was saturated.

44

However, the oxygen present in the soil above 61.0

cm was not sufficient to convert large quantities of reduced nitrogen to nitrates, or the organic nitrogen was in forms that were not readily oxidizable.

Nearly twice as much total nitrogen was detected in the filtrate at 61.0 cm as was in the applied effluent during the tenth week.

Similar nitrate results were obtained for the C part of the

C-2F5D treatment but increases in the nitrate content of the filtrate occurred during the final five weeks when the 2F5D treatment was employed.

During this five-week period, the greatest amounts of nitrate were detected in the filtrate on the first of the two consecutive days of irrigation, probably the result of nitrification occurring during the drying period.

The losses of organic nitrogen and ammonium, probably by nitrification, indicate the presence of an aerobic environment in the upper 61.0 cm of soil while the loss of total nitrogen indicates that denitrification, or possibly immobilization and reduction to ammonia, occurred within the same soil layer.

An anaerobic or nearly anaerobic layer may have been established immediately above the

61.0 cm depth due to the impedance of drainage by the change in soil texture at that level.

Soil moisture tensions increased from 35.0 to

42.6 cm of water during the first three days of the drying periods but leveled off during the fourth and fifth days.

This change probably indicates that drainage of most gravitational water from the soil columns occurred within three days after cessation of irrigation.

Drainage from the lower part of the soil columns during the five day

45 drying periods is indicated by the tensiometer data but the extent of reaeration cannot be ascertained since no nitrogen data was obtained at the 243.8 cm depth during the final five weeks.

In the alternate day treatment, the nitrate content of the filtrate increased on the third day of the test, rather than decreased as in the continuous flooding treatment.

This was attributed to the partial oxidation of nitrogen compounds deposited in the surface soil layer during the first day of irrigation.

Leaching of nitrate from this layer after one day of drying, in addition to some nitrification occurring near the surface, probably caused the higher nitrate content at 61.0 cm depth.

A decrease in the nitrate content of the filtrate during the final weeks was either the result of lower quantities of total nitrogen being applied or the presence of a more anaerobic environment in the surface 61.0 cm soil layer of the 1F1D columns.

Tensioneter data indicate partial drainage of these columns occurred during the one-day drying portion of the irrigation cycles.

Higher nitrate contents in the filtrate at the bottom of the soil columns indicate the presence of a partially aerobic environment between 61.0

and 243.8 cm.

Fven larger quantities of nitrate in the filtrate at 61,0 cm depth of the 1F3D columns than those found in the 1F1D columns suggest the presence of a more aerobic environment than in the latter treatment.

Total nitrogen losses were comparable in both cases.

The increased nitrate content of the 1F3D filtrate was probably the result of a partal1y aerobic environment which limited the denitrification process in the upper soil layers.

Increases in soil moisture tension indicate

that some drainage of gravitational water was still occurring at the bottom of the soil columns during the third day of the drying period.

46

The extent of reaeration in the lower part of the 1F3D columns is unknown.

Trace Metal Filtration

Trace element removal from sewage effluent was studied by applying four different flooding treatments to lysimeter columns and determining the trace metal content of the applied effluent, filtrate, soil and vegetation removed.

The quantities of the various trace metals applied to each treatment are summarized in Table 4 for the 11 week test.

Strontium was applied in the largest concentrations, from 0.9 to 1.3 mg/i.

Average iron applications ranged from 0.37 to 0.42 mg/i for the four treatments.

Nearly equal amounts of zinc, copper and lead were applied but the average concentrations were consistently less than 0.1 mg/i.

From 0.02 to 0.03 mg/i of manganese was applied to the lysimeter columns during the study.

Slightly lower quantities of cadmium were detected in the applied effluent.

No detectable amounts of cobalt or chromium were applied to the columns during the 11 week test.

The sodium acetate-extractable trace elements found in the surface 61 cm of soil, before and after irrigation, are listed in Table

5.

I general, sodium acetate-extractable iron, lead and cadmium were decreased by the irrigation treatments while manganese increased, especially for the continuously flooded treatment.

Little change was observed in the nickel, copper, zinc or strontium levels.

No

meaningful amounts of sodium acetate-extractable chromium or cobalt were detected in the soil samples.

Prior to irrigation, the largest

49 concentrations of nickel, copper, zinc, lead and strontium in the surface 30.5 cm layer of soil were located in the 22.9 to 30.5 cm horizon.

The accumulation of these metals at this depth was probably the result of leaching of water soluble materials with deionized water used to irrigate the grass prior to effluent application.

The concentrations of copper, zinc and lead at this depth dissipated during the course of the study by deeper leaching or by conversion to some form not amenable to sodium acetate extraction.

The amounts of trace metals removed from the filtering system in the bermudagrass clippings are listed in Table 6.

The densities of grass on the soil columns for each treatment are apparent from the weights of grass removed from each treatment during the study.

The most dense grass cover was established on the lFlD columns with a slightly sparser cover on the 1F1D-1F3D columns.

A very sparse grass cover appeared on the C-2F5D columns during the final two weeks of the test but no bermudagrass became established on the C columns during the test.

Iron and manganese were removed in the largest quantities in the grass clippings.

Small amounts of copper, zinc and cadmium were also removed in the grass but no chromium, nickel, lead, cobalt or strontium was detected in the clippings.

Iron

The removal of iron from the applied effluent apparently was accomplished within the surface 7.6 cm of soil, in the lysimeter

51 columns, for the four treatments employed (Fig. 7).

Of 880.5,

1348.4, 1193.3 and 1228.1 mg of iron applied in the C, C-2F5D, 1F1D and 1F1D-1F3D treatments, respectively, only 2.2, 4.0, 5.4 and 2.0 mg were detected in the filtrate at the 7.6

cm depth.

The only iron detected in the filtrate at 15.2 cm depth was during the eighth week for the 1F1D treatment.

At least half, and possibly all, of the 3.3

mg at this point originated from the soil material,

Prior to irrigation, the top 7.6

cm soil layer of each treatment set of columns contained an estimated 31.5 mg of sodium acetateextractable iron (Table 5).

Values of 23.4, 21.9 and 3.7 mg were detected in successively deeper 7.6

cm layers.

Following 11 weeks of irrigation, 7.2 mg of sodium acetate-extractable iron remained in the upper 7.6 cm layer of soil in the continuously flooded columns.

Smaller and decreasing amounts, 6.8, 5.0 and 0.4 mg, were found in the soils of the three deeper layers for the C treatment.

In the C-2F5D treatment, 6.2 mg of sodium acetate-extractable iron were retained in the 0.0 to 7.6 cm soil layer while only 1.0 mg remained in this layer in the 1F1D columns.

In the 1F1D columns an increase In iron content was observed with depth, in the upper 30.5 cm of soil.

The columns receiving the lFlD-1F3D treatment contained 3,7 mg of sodium acetateextractable iron in the surface 7.6 cm layer after 11 weeks of effluent irrigation.

One, 3.2 and 2.6 mg were found in the successively deeper

7.6 cm soil layers.

The average quantities of iron removed from the filtering system by the common bermudagrass were 0,46 mg per gram of dry grass in the

53

C-2F5D treatment, 0.28 mglg in the

1F1D treatment and 0.46 mg/g in the 1F1D-1F3D treatment (Table 6).

The minor amounts of iron in the water at the 7.6 cm depth throughout the study, the presence of only small quantities of sodium acetate-extractable iron in the 0.0 to 7.6 cm soil layer at the termination of the test and the low amounts of iron removed in the grass clippings suggest that the applied iron must have been deposited near the soil surface in some form not amenable to sodium acetate extraction.

Some iron may have been assimilated by the soil microorganisms, or formed organic complexes, but most probably precipitated in inorganic forms.

Loss of sodium acetate-extractable iron from the surface 61.0 cm of soil, combined with the failure to detect more than minor amounts of iron being translocated downward, suggest that this iron was converted to some other form.

The bulk of the iron was probably deposited as oxides of iron and as ferric hydroxide, a compound reported to be formed in neutral or alkaline soils (Olson,

1965).

The sodium acetate-extractable iron remaining in the soil at the termination of the test is indicative of the aeration of the soil at that time.

The drainage and aeration status of the lysimeter columns was presented in a previous section.

The largest amounts of iron detected in the surface layer were in treatments C and C-2F5D with lesser quantities in the better-aerated 1F1D and lFlD-1F3D treatments.

This same trend is evident in the continuously flooded columns in which the amount of sodium acetate-extractable iron decreased with

54 depth.

The decreasing amounts of iron with depth suggest possible saturation at the surface but partial drainage and oxidation of iron in the 15.2 to 22.9 and 22.9 to 30.5 cm soil layers.

The increase in sodium acetate-extractable iron in the second 30.5 cm soil layer could have been caused by poor drainage induced by the finer material in the

61 to 91 cm layer.

The small amounts of sodium acetate-extractable iron in the surface layer are indicative of the favorable aeration status of the

1F1D and lFlD-1F3D treatments.

Increasing quantities of iron with depth in the 1F1D columns could be attributed to the better drainage and aeration of the surface than occurred at successively lower layers during the one day drying period in each cycle.

The quantities of iron taken up by the grass were small in view of the amounts applied in the effluent and did not account for more than 2.6 percent of the iron removed from the effluent.

On the basis of the filtrate results, none of the treatments employed would be preferential to any other for iron removal from sewage effluent.

However, the lower quantities of iron retained in sodium acetate-extractable form and the maximum removal of iron in the grass make the 1F3D treatment more favorable than any other treatment.

Also more effluent was reclaimed by the 1F3D columns than in any of the other three treatments used during the final five weeks.

Manganese

Nearly 70 mg of manganese were applied to the continuously flooded columns, during the 11 week test, with 21.3 mg being detected

55 in the filtrate at the 7.6 cm depth (Fig.

8).

Even larger quantities of manganese were detected in the water at the 15.2 and 22.9

cm depths, especially during the second and third weeks.

Nearly 64 mg of manganese were detected in the filtrate at the 61.0 cm depth during the first six weeks of continuous inundation with the bulk being measured during the third and fourth weeks.

Only 8.7 and 2.5 mg were detected at the 121.9

and 182.9 cm depths, respectively.

The observation of 3.0 and 4.5

mg of manganese in the filtrate samples from the 121.9 cm depth during the fifth and sixth weeks, respectively, possibly marked the penetration of manganese to this depth.

Samples were not obtained at this level during the final five weeks, therefore, it is not certain whether this trend continued.

No appreciable quantities were detected at 243.8

cm, and none at all after the third week.

During the continuously flooded period of the C-2F5D treatment, movement of manganese similar to that discussed above was observed.

Of 113.8 mg of manganese applied in the effluent, 43.3 mg were detected in the filtrate at 7.6 cm depth.

The manganese content of the water was 111.1, 152.5 and 190.3 mg at the 15.2, 22.9 and 30.5 cm depths, respectively, during the 11 weeks.

Little manganese movement was observed at the 182.9 and 243.8 cm depths.

Once again, the bulk of the manganese translocation in the upper soil layers occurred in the first three weeks with deeper movement occurring during the fourth, fifth and sixth weeks.

During the latter five weeks of the C-2F5D treatment when the 2F5D irrigation pattern was employed, increased manganese movement was observed at all four sampling depths in the

57 upper 30.5 cm of soil,

Very small but increasing amounts of manganese were observed at the 243.8 cm depth during the final five weeks.

In the 1F1D treatment columns, little manganese was detected in the filtrate samples extracted below 7.6 cm depth but 75,4 mg were observed at the 7.6 cm level, effluent.

More than 95 mg were applied in the

As in the continuously flooded columns, the largest quantities of manganese at the 121.9 cm depth were observed during the fifth and sixth weeks.

The same manganese filtration pattern was observed for the

1F1D-1F3D treatment as for the 1F1D method.

However, slightly less manganese was observed at 7,6 cm during the final five weeks of the

1F3D test than in the 1F1D treatment during the same period.

During the entire 11 weeks, 45.4 mg were detected at the 7.6

cm depth compared to 106.1 mg in the applied effluent.

No sodium acetate-extractable manganese was found in the upper

61 cm of soil prior to effluent irrigation (Table 5).

At the termination of the test the greatest amounts were found in the continuously flooded soils.

The amounts measured in successively deeper 7.6 cm soil layers were 17.5, 29.9, 42,3 and 38.7 mg.

In the other three treatments, these same soil layers contained less than five mg of sodium acetate-extractable manganese and a general decrease with depth was observed.

Although only a trace of manganese was detected in the grass sample from the C-2F5D treatment, more substantial quantities were removed from the 1F1D and 1F1D-1F3D columns.

The average manganese

58 contents of the grass samples from these respective treatments were

0.21 and 0.22 mg per gram of dry grass.

Manganese uptake by the grass accounted for 22 percent of the amount applied in the effluent for the

1F1D treatment and 15 percent of the applied amount in the sparser grass cover of the 1F1D-1F3D treatment.

The amounts of manganese removed from the effluent, and the soil depth at which filtration occurred, were not as apparent as those for iron.

Part, or all, of the manganese detected in the filtrate of the C columns may have originated from the soil material.

Two factors substantiate this point.

First, approximately 1.5 to 9.0 mg of manganese were detected in the filtrate samples at depths of 7.6 and

15.2 cm, from the second through the fifth weeks, thereafter decreasing to less than 1.2 mg per week, even though the amounts of manganese applied were approximately the same as during the earlier weeks.

Second, greater quantities of manganese were consistently detected at the 15.2 cm depth than the total amount that had penetrated the

7.6 cm depth in the previous weeks.

Even further increases were observed in the filtrate at the 22.9 cm depth but less was observed at

30.5 cm, indicating the deposition of some of the translocated manganese between 22.9 and 30.5 cm in the continuously flooded columns.

Manganese mobilization, from precipitated iron and other oxides, under anaerobic conditions in the presence of fermenting plant materials was reported by Ng and Bloomfield (1962) and could have been the cause of manganese moving through the soils of the continuously flooded columns.

The method for manganese mobilization may have been the same during the C part of the C-2F5D treatment but this process by itself

does not account for the large amounts of manganese translocatjon that occurred during the better-aerated 2F5D phase of the treatment.

Neither will the high infiltration rates alone account for the large

59 quantities of manganese movement through the soil since no manganese translocatjon was observed below 7,6 cm of the 1F3D treatment which maintained even greater infiltration rates.

Manganese movement was apparently the result of a combination of the high infiltration rates and a longer period of saturation, two days versus one day, causing the reduction and movement of larger quantities of manganese in the soil material.

In the 1F1D and 1F1D-1F3D treatments manganese movement was restricted to the upper 15 cm of soil because of the better aeration provided by the alternate day dry periods and less reduction of manganese compounds by the shorter periods of saturation.

Reduction of manganese compounds in the presence of fermenting organic material is also substantiated by the results of the soil extractions.

The largest quantities of sodium acetate-extractable manganese were found in the soils of the continuously flooded columns.

The increase with depth, to a maximum value of 42.3 mg in the 15.2 to

22.9 cm layer, indicates that extensive leaching occurred in the anaerobic environment at the top of the columns.

The greater deposition of manganese occurring in the lower part of the top 30.5 cm soil layer is apparently the result of the improved drainage and aeration of this soil layer.

The lesser quantities of sodium acetate-extractable manganese in the lysimeter columns receiving intermittent irrigation are in

60 agreement with the manganese mobilization results of Ng and Bloomfield

(ibid).

The decreased manganese content with depth apparently resulted from improved aeration of the soil layers between

7.6 and 30.5 cm.

The manganese detected in the filtrate at the 121.9 cm depth in all treatments during the fifth and sixth weeks was probably mobilized in the upper soil layers and transported downward in the percolating water.

This could have been possible in the C and

C-2F5D treatments, since large quantities of manganese were detected in the filtrate at all higher sampling layers, but could riot have been the source of manganese in the 1F1D and 1F1D-1F3D treatments since only minor quantities penetrated the next higher sampling layer.

Therefore, in the 1F1D and 1F1D-1F3D treatments, and possibly in the C and C-2F5D treatments, the manganese must have been derived from the soil material between 61.0 and 121.9 cm.

Either manganese originated from water soluble material in the soil or was mobilized in the possibly anaerobic environment of the fine-textured 61 to 91 cm soil layer.

On the basis of the experimental results discussed above, the irrigation pattern that would allow the maximum removal of manganese from the applied effluent would be one of less than one day of flooding followed by a drying and reaeration period.

A dense common bermudagrass cover would facilitate removal of manganese from the soil system provided the manganese is not leached below the root zone.

Of the cyclic irrigation patterns used in this study, the 1F3D treatment produced the best results from a management viewpoint.

A greater volume of effluent was treated by this method than in the 1F1D treatment while comparable renovation results were attained.

61

Chromium form.

The only chromium ion included in this study was the hexavalent

None was detected in any of the effluent, filtrate or grass samples and only 0.3 mg was found in one soil sample, the 22.9 to 30.5

cm layer of the continuously flooded columns at the termination of the study.

This value was believed to be an error attributable to contamination of equipment.

Nickel

Only the continuous flooding and 1F1D treatments can be evaluated for nickel filtration since nickel was only found in samples of effluent applied to the lysimeter columns during the second, third, fifth and sixth weeks (Fig. 9).

No translocation of nickel was detected during the final five weeks for any of the four treatments employed during this period.

Nearly identical nickel translocation patterns were observed in both sets of three columns receiving continuous irrigation.

Of

28.3 and 36.8 mg of nickel applied to the C and C-2F5D columns, respectively, only 3.2 mg were detected at the 7.6 cm depth in each treatment, The nickel contents decreased to 2.6 and 2.8 mg at the

152 cm depth in the C and C-2F5D treatments before increasing to

10.9 and 11.9 mg at the 22.9 cm depth.

No nickel was detected in the filtrate at 30.5 cm in either treatment but 7.3 and 1.1 mg were detected at the 61.0 and 121.9 cm depths of the C treatment.

The filtrate at these respective depths of the C-2F5D columns contained

9.8 and 2.4 mg of nickel.

Of the 43.8 mg of nickel applied to the 1F1D and 1F1D-1F3D columns only 10,1 mg were detected in the filtrate at 7.6

cm.

Only minor quantities of nickel were detected in the filtrate samples obtained below the 7.6 cm depth.

63

The only sodium acetate-extractable nickel in the upper 61 cm of soil, prior to irrigation, was 0,5 mg observed in the 22.9 to 30.5

cm soil layer (Table 5).

Following 11 weeks of continuous irrigation, minor amounts of nickel were detected throughout the upper 61 cm of soil.

Minor quantities were also detected at random depths within the surface 61 cm of the C-2F5D and 1F1D treatments.

No sodium acetateextractable nickel was found in the 1F1D-1F3D treatment soil at the termination of the study.

Nickel was not removed in any of the grass clippings from any of the treatments employed.

Like manganese, much of the nickel moving downward through the continuously flooded columns was probably derived from soil material, being mobilized from oxides in the potentially anaerobic soil environments.

The larger quantities of nickel at the 22.9 cm depth, than the amounts that penetrated the 15.2 cm depth, and the observation of 4.1

mg of nickel in the filtrate at the 22.9 cm depth during the first week of irrigation are evidence of the translocated nickel being derived from soil material rather than from the applied effluent.

Not more than 5.4 pig of the applied nickel, or 8.3 percent, could have penetrated beyond 15.2 cm in the continuously flooded columns.

The absence of nickel movement below 121.9 cm was attributed to the rapid drainage of

64 water from the sands and fine gravels located in the lower half of the soil columns.

Smaller amounts of nickel translocation in the 1F1D columns were attributed to the better drainage and aeration of this soil system.

In the 1F1D treatment more than 77 percent of the applied nickel was removed from the filtrate within the surface 7.6 cm layer of soil and all applied nickel was eliminated from the filtrate above the 15.2 cm depth.

The small quantities of sodium acetate-extractable nickel

detected in thesoils at the termination of the

test, the failure to detect any appreciable amounts being leached down through the columns and the absence of nickel in the grass samples indicate that most of the nickel was deposited in some form not amenable to sodium acetate extraction.

More sodium acetate-extractable nickel was found in the continuously flooded columns than in the lFlD columns, possibly the result of the reducing conditions present in this treatment.

Of the two treatments employed for nickel filtration, the more favorable results were obtained by the 1F1D treatment.

Nickel penetration in the filtrate was restricted to the upper 15.2 cm of the soil columns and deeper movement was limited in this treatment.

Complete removal of nickel from the applied effluent was also attained in the

C treatment but deeper translocation was observed.

Copper

Large, but irregular, quantities of copper were applied in the effluent for all treatments.

However, only relatively small amounts penetrated beyond the upper soil layer (Fig. 10).

a- a-

66

Of 195.5 mg applied to the continuously flooded columns, only

11.2 mg, or 5.7 percent, were detected at 7.6 cm depth.

Similarly small copper contents were noted for all sampled depths of the

C treatment.

During the 11 week test, 21.1 mg of copper were detected at the bottom of the soil columns but approximately two-thirds of this was observed during the first four weeks.

After that time the weekly copper contents ranged from 0.9 to 1.6

mg.

Similar results were obtained for the C part of the C-2F5D treatment.

However, during the latter part of this treatment increasing quantities of copper were detected in the filtrate at the 7.6,

15.2, 22.9 and 30.5 cm depths.

Of the 253.5 mg applied in the effluent during the 11 weeks, 21.3 mg were detected in the filtrate at the

243.8 cm depth.

During the 2F5D phase of the treatment, 99.6, 13.1,

10.7, 8.2 and 11.1 mg were detected in the applied effluent and at the

7.6, 22.9, 30.5 and 243.8 cm depths, respectively.

Nearly ten percent of the 239.6 mg of copper applied in the

1F1D treatment was detected in the filtrate at the 15.2 cm depth while

16.2 percent appeared at 7.6 cm.

More than 22 mg were detected at the

243.8 cm depth in weekly concentrations ranging from 0.6 to 3.7 mg.

Under the lFlD-1F3D treatment similar reductions in the copper content were observed at the 7.6 cm depth.

Even better filtration, than in the 1F1D treatment, was attained above 15.2 cm depth where only

15.6 mg were detected in the filtrate.

The copper content of the water increased to 22.2 and 29.8 mg at the 22.9 and 30.5 cm depths, respectively.

Nearly 20 mg were observed in the filtrate at the 243.8 cm depth.

The sodium aCetate-extractable copper content of the 0.0 to

7.6, 7.6 to 15.2, 15.2 to 22.9 and 30.5 to 61.0 cm layers of soil increased during the course of the study in all treatments (Table 5),

67

However, the quantity of copper in the 22.9

to 30.5 cm soil layer was consistently reduced.

Initially, this soil layer contained ten times more sodium acetate-extractable copper, 6.7 mg, than found in any other 7.6 cm soil layer in the surface 30.5

cm.

Following 11 weeks of irrigation, only the 7.6 to 15.2 cm layer of the C-2F5D treatment contained more than 1.8 mg of sodium acetate-extractable copper.

The

C-2F5D treatment contained the largest quantities of sodium acetateextractable copper at the end of the study.

Only 0.2 and 0.5 mg of copper were removed from the filtering systems in the grass clippings from the 1F1D and 1F1D-1F3D treatments, respectively (Table 6).

Although some copper was observed in samples from all sampling layers within the columns, the concentrations of copper passing 243.8

cm was insignificant, even by United States Public Health Service (1962) drinking water standards (1. rng/l).

In the C treatment 21.2 mg were detected in an estimated 2212.8 liters of infiltrated water, an average of less than 0.01 mg/i.

The copper concentrations in the percolate at

243.8 cm were 21.3 mg in 3228.0 liters, 22.1 mg in 2713.5 liters and

19.9 mg in 2829.7 liters for the C-2F5D, 1F1D and 1F1D-1F3D treatments, respectively.

The average copper concentrations in the percolate, for these respective treatments, were 0.007, 0.008 and 0.007 mg/l.

Some of the results indicate that part of the copper moving through the lysimeter columns was derived from soil material.

For

example, larger quantities were observed in the filtrate at the 22.9

68 and 30.5 cm depths than in the filtrate at the 15.2 cm depth of the lFlD-1F3D treatment.

In freely drained soils, copper is reported to be sorbed by organic materials (Mitchell, 1964).

The dissolved and suspended organic fractions in the percolating water may have been the vehicle by which copper was moved downward through the soil material in the intermittent irrigation treatments.

Less copper translocation in the C treatment and the C part of the C-2F5D treatment may have been the result of sealing of the soil surface in these treatments, thereby effecting the removal of most organic materials from the effluent at the soil surface, thus reducing the organic fraction for copper transport.

Laverty et al.

(1961) have reported a slime layer to be formed on the soil surface within a few days when continuous irrigation was applied.

This organic mat not only controlled the infiltration rate but also the intake of solid particles.

Under the anaerobic treatments, copper mobilization, like that reported for several other elements (Ng and Bloomfield, 1962), could have accounted for part of the movement of copper through the soil.

The low sodium acetate-extractable copper concentrations in the surface soil layers indicate a low retention of copper by the cation exchange sites.

Not more than two milligrams of sodium acetateextractable copper were retained in the surface 7.6 cm of soil for any of the treatments, either on the cation exchange sites or as extractable precipitates.

This could also be the cause for the decrease in the copper concentrations in the 22.9 to 30.5 cm layer during the course

69 of the study since the same soil material was used for the upper

30.5 cm of the lysimeter columns.

The extremely small amounts of copper removed in the grass clippings indicate that little maintenance of the filtering capacity of the soil systems can be expected by grass uptake of copper ions,

Although copper is reported to be assimilated by microorganisms

(Starkey, 1955) no appreciable loss in copper can be accomplished by this method since the metal will be returned to the soil upon death and lysing of the organisms.

The presence of organic materials in the filtrate and the failure of biological assimilation to remove any appreciable quantity of copper from the filtering system limit the capacity of the soil for removing copper from the applied effluent.

Even though copper concentrations of less than 0.01 mg/i were estimated to be in the filtrate at the bottom of the soil columns throughout the study, there is an apparent danger of soil saturation with copper in a relatively short period of time.

From the results of this study, it is difficult to determine which of the four treatments produced the most desirable copper filtration.

More than 90 percent of the applied copper was removed within the surface 15.2 cm of soil for all treatments.

However, minor quantities penetrated the entire length of the soil columns.

Slightly less copper movement was observed in the C treatment, possibly the result of more complete removal of the organic materials by the sealed surface layer.

Even though larger quantities of copper were detected in the percolates of the intermittently flooded columns, lower average

70 concentrations were attained in view of the greater quantities of water passing through the soil columns.

Zinc

Zinc, like copper, was applied in the effluent in large but irregular quantities, was detected in the filtrate samples from all depths for all treatments and was not present in undesirable quantities in the percolate at the bottom of the lysimeter columns.

Under continuous irrigation, 79.0 and 50.8 mg were detected in filtrate samples from 7.6 and 15.2

cm, respectively, compared to 196.3

mg in the applied effluent (Fig. 11).

During the 11 weeks, 20.2 mg of zinc were detected in the percolate at the 243.8 cm depth but 12.7

mg of this amount was found during the first week.

No zinc was found to penetrate the 121.9 cm depth during the first week.

Zinc translocation in the C columns occurred primarily during the first four weeks of irrigation.

Less than two mg per week were detected in the filtrate at depths below 22.9 cm after the initial four-week period.

Similar results for zinc penetration were obtained from the C phase of the C-2F5D treatment.

Increased zinc translocation was observed during the latter portion of this treatment when the 2F5D cycles were employed.

During the final three weeks of the test, from

15 to 30 percent of the 94.5 mg of zinc applied was observed at the

243.8 cm depth but the average weekly concentrations were only 0.013,

0.009 and 0.005 mg/l.

In the 1F1D treatment 254.6 mg of zinc were applied to the soil columns during the 11 week test.

During the same period 55.2, 79.8,

57.3 and 42.5 mg were detected in the filtrate at 7.6,

15.2, 22.9 and

72

30.5 cm depths, respectively.

Twenty-three mg were observed in the filtrate at the 243.8

cm depth.

The amount of zinc in the filtrate generally declined during the course of the Study at all sampled depths of the soil columns, with the exception of the final week at the 15.2,

22.9 and 30.5 cm depths.

A similar zinc translocation pattern is evident for the 1F1D part of the 1F1D-1F3D treatment.

During the latter five weeks, when the 1F3D cycles were used, the effects were similar to those obtained with the 2F5D treatment but total zinc movement was greater, especially during the final three weeks.

More zinc translocation was observed in this treatment than for any other even though more zinc was applied in the 1F1D and C-2F5D treatments.

Approximately 43 mg of zinc were detected in the filtrate at the 243.8 cm depth during the test, compared to 247.9 mg in the applied effluent.

Of the 43 mg, 9.0, 7.2 and

14.6 mg were found during the first, tenth and eleventh weeks, respectively.

The quantities of sodium acetate-extractable zinc in successive

7.6 cm soil layers, in the surface 30.5 cm of soil, were 1.9, 0.8, 0.3

and 2.1 mg prior to effluent irrigation (Table 5).

In most instances, smaller amounts were detected following treatment.

No pattern was apparent for sodium acetate-extractable zinc remaining in the soil following the various treatments.

An estimated 0.1 mg of zinc was removed in 1.7 mg of dry grass from the C-2F5D columns for an average of 59 iig/g (Table 6).

Larger quantities, but smaller concentrations, were removed in the common

73 bermudagrass clippings from the 1F1D and 1F1D-1F3D treatments.

An estimated 2.2 mg of zinc was removed from the 1F1D columns and 0.7 mg from the 1F1D-1F3D columns for averages of 22 and 10 i-tg/g, respectively.

Zinc movement within the lysimeter columns was greatest during weeks when infiltration rates were high.

During the first week, when infiltration rates were maximum, zinc applications ranged from 11.6 to

13.8 mg and the quantities detected in the filtrate at the 7.6 cm depth ranged from 12.1 to 17.7 mg, for the different treatments.

During the second week, as infiltration rates decreased and zinc applications increased, the quantities of zinc penetrating the top four sampling layers decreased from the previous week, for all treatments.

Infiltration rates and zinc concentrations in the filtrate, at the four sampling depths in the surface 30.5 cm, increased during the third week then declined in the fourth and fifth weeks.

Zinc movement was limited after that time in the C and 1F1D treatments.

However, zinc movement was considerable in the surface 30.5 cm of the 2F5D treatments during the final three weeks when the infiltration rates increased to their highest values since the first week.

More zinc translocation was observed at the 243.8 cm depth of the 1F3D treatment during the tenth and eleventh weeks than ip the other treatments.

The 1F3D columns had the greatest infiltration rates during the final two weeks.

More evidence favoring the relationship between zinc movement in the soil and infiltration, rather than the quantity of zinc applied, can be seen from the data obtained during several weeks at the middle of the test period when relatively large amounts of zinc were applied but only minor quantities

were detected in the filtrate at any depth.

Infiltration rates were generally low during this period.

74

In the C treatment, more than 75 percent of the applied amount of zinc was removed in the upper 15.2 cm of soil during the ii weeks.

Nearly half of the 99.5 mg of zinc applied during the first four weeks was detected in the filtrate at the 30.5

cm depth but during the final seven weeks only 7.3 mg of 96.8 mg applied was found at this depth.

The average zinc concentration at the bottom of the soil columns in the C treatment was less than 0.01 mg/i, an insignificant quantity in comparison to the USPHS drinking water standard (1962) of less than

5. mg/i.

Even though zinc movement increased in the 2F5D treatment, less than 23 percent of the applied amount was detected in the filtrate at the 30.5 cm depth.

Only ten percent of the zinc applied to the 1F3D columns in the final six weeks reached the 30.5 cm depth.

Under the lFlD treatment, zinc filtration results were more comparable to those obtained in the C treatment than the results from the 2F5D or 1F3D treatments, This was attributed to the lower infiltration rates under the lFlD treatment.

Some of the zinc translocated downward in the soil columns was derived from the soil media and could have been released by fermenting organic materials.

The necessary anaerobic conditions could have been provided by high infiltration rates causing the rapid saturation of the upper soil layers for short periods of time.

Less zinc movement in the soil layers below 30.5 cm was attributed to the more rapid drainage of the coarser-textured materials.

75

On the basis of these results, infiltration rates must be considered an important factor for maximum zinc removal from the effluent.

Apparently saturation of the soil for even short periods of time is sufficient to allow the translocation of zinc through the soil.

Of the four treatments employed in this study, the least zinc movement was detected in the C and 1F1D treatments, primarily because of the low infiltration rates in these two treatments and because of the slightly aerobic environments in the lower part of the soil columns.

However, reductions in the zinc concentrations to desirable levels were achieved in all treatments.

Bermudagrass uptake of zinc was minimal and could not be expected to maintain the zinc filtration capacity of the soil.

Lead

Large quantities of lead were found in the effluent applied to the soil columns during the first, third, fourth, eighth and eleventh weeks (Fig. 12).

Of 172.6 mg applied to the C columns only 1.5 mg were detected in the filtrate at the 7.6 cm depth.

Five and 6.6 mg were observed in the filtrate at the 15.2 and 22.9 cm depths, respectively.

Only 1.6 and 1.5 mg were found at the 30.5 and 61.0 cm depths, respectively, but 14.8 mg were detected at the 121.9 cm depth.

No lead was found in the filtrate at the 182.9 cm depth and less than two mg were observed at

243.8 cm.

Lead was also found in the filtrate during the C part of the

C-2F5D treatment.

Exactly the same lead detection pattern was observed for both continuously flooded treatments, with only slightly different

77 amounts of lead detected in the two treatments.

Of the 221.9 mg of lead applied to the C-2F5D columns Only 1.7 mg appeared at the 7.6 cm depth.

Less than three mg of lead were detected at the 7.6 cm level of the 1F1D columns but 9.2

mg were detected at the 15.2

cm depth.

None of the applied 162.7 mg penetrated farther than 22,9 cm since no lead was detected at this depth.

The lead observed in the filtrate at 30.5

and 121.9 cm originated in the soil material.

In the 1F1D-1F3D treatment, 197.5 mg of lead were applied in the sewage effluent but only 2.5 and 1.3 mg were detected at the 7.6

and 15.2 cm depths, respectively.

Lead detected in the filtrate at the

22.9, 30.5 and 121.9 cm depths apparently was derived from soil material.

Before effluent irrigation, no sodium acetate-extractable lead was found in the surface 15.2 cm of soil and only 0.8 mg was found in the 15.2 to 22,9 cm layer of each treatment set of columns (Table 5).

The 22.9 to 30.5 cm layer and the 30.5 to 61.0 cm layer contained 55.2

and 23.9 mg, respectively.

The four different irrigation treatments produced erratic changes in the lead content of the soil layers sampled.

In the continuously flooded columns lead was found in four of the five soil layers, the maximum quantity being in the surface 7.6 cm of soil.

Under the C-2F5D treatment, 2.8 and 5.3 mg of sodium acetateextractable lead were detected in the 7.6 to 15,2 and 15.2 to 22.9 cm soil layers, respectively.

None was found in the surface 7.6 cm or between 22.9 and 61.0 cm.

Under the 1F1D treatment only 0.4 mg was detected in the 0.0 to 7.6 cm layer.

In the 1F1D-1F3D treatment 1.2 and

78

3.5 mg of sodium acetate-extractable lead were detected in the

7.6 to

15.2 and 30.5 to 61.0 cm soil layers, respectively.

No lead was detected in the other layers sampled.

No measurable quantities of lead were detected in the grass clippings removed from the C-2F5D, lFlD and 1F1D-1F3D treatments.

More than 98 percent of the applied lead was removed from the filtrate in the surface 7.6 cm of soil for all treatments.

However, in the C treatment and the C part of the C-2F5D treatment lead was mobilized and translocated down through the entire soil columns.

Less lead movement was observed in the more-aerobic 1F1D and 1F1D-1F3D treatments.

No translocation of lead was observed in the 2F5D treatment.

These results agree with the lead mobilization results of Ng and Bloomfield (1962) who reported the mobilization of lead in the presence of fermenting organic materials.

The lead in the filtrate at the 121.9 cm depth probably originated between the 61 and 91 cm depths.

The larger amounts of silt and clay in this layer could have restricted the drainage of water and caused the resultant anaerobic conditions necessary for lead mobilization.

Deposition of lead in the soil below

121.9 cm was to be expected since the coarser material throughout the lower half of the soil columns allowed partial drainage and reaeration of the soil.

The larger amounts of sodium acetate-extractable lead in the C columns at the termination of the study, than in the intermittently flooded columns, were the result of the lead reducing conditions of the soil environment.

On the basis of nitrogen data, the upper soil layers of the C columns were estimated to be at least partially anaerobic.

79

The decrease in extractable lead in the 22.9 to 61.0 cm soil layer was probably due to leaching of lead to lower levels and to lead conversion to nonextractable forms.

The results of this study indicate that lead translocation can be minimized by maintaining an aerobic soil environment.

Excellent lead filtration from sewage effluent was accomplished by all treatments applied.

Cadmium

Cadmium was found in samples of the effluent applied to the soil columns during the third, fourth, sixth and seventh weeks but none was detected in the filtrate at any of the sampled depths (Fig.

13).

Even during the seventh week when more than 83 mg of cadmium were applied to the 1F1D columns, in 354.6 liters of effluent, none was found to be translocated through the surface 7.6 cm of soil.

Sodium acetate-extractable cadmium in the surface 30.5 cm of soil, prior to irrigation, amounted to 1.0 mg and in the second 30.5

cm layer to 0.9 mg (Table 5).

After the test, traces of cadmium remained in the soils of the continuously flooded columns but none was detected in the columns receiving any type of intermittent irrigation.

Small amounts of cadmium were removed in the grass clippings

(Table 6).

One of the clippings taken from the 1F1D columns contained

This value was all 3.6 mg of cadmium removed from this treatment.

believed to be an error caused by contamination since no later samples from this set of columns contained any cadmium.

The results indicate complete cadmium removal from the water

81 at, or near, the soil surface for all treatments and the absence of any translocation of cadmium within the columns.

Cobalt

Although the cobalt content of the applied effluent was below detectable levels, some movement in the filtrate was noted (Fig. 14).

Translocation was restricted to the first five weeks with the exception of 7.2 mg in the filtrate at the 7.6 cm depth of the lFlD columns during the final two weeks and 3.1 mg at 22.9 cm in the C columns during the final week.

No cobalt was removed in the grass clippings and, with the exception of the 7.6 to 15.2 cm layer of the 1F1D treatment, none was detected in the soil samples by sodium acetate exchange.

Strontium

Alkaline soil, such as that contained in the lysimeter columns, is ineffective in removing strontium from percolating water (Fig. 15).

The amounts of strontium applied were considerably greater than for any of the other elements tested, with weekly applications of up to 1650 mg.

A total of 3130.4 mg of strontium were applied to the

C columns during the 11 weeks of irrigation.

The minimum amount detected in the filtrate at any sampled depth was 1283.7 rug at the

30.5 cm level.

The strontium content of the filtrate increased to

2495.4 mg at the 243.8 cm depth.

The strontium content of the applied effluent in the C-2F5D treatment was 4461.9 rug.

A minimum amount of 1913.6 mg was detected

84 at the 22.9 cm depth and 4299.8 mg were detected at the bottom of the soil columns.

In the 1F1D treatment 2593.6 mg were applied in the effluent and 2677.7 mg were detected in the filtrate at the 243.8 cm depth.

The minimum strontium content observed at any sampled depth was 1522.1 mg at the 15.2 cm depth.

A strontium content of 2295.0 mg was detected at the bottom of the lFlD-lF3]J columns compared to 2745.2 mg applied in the sewage effluent.

The lowest strontium content observed within the columns was

1896.4 mg at the 22.9 cm depth.

The sodium acetate-extractable strontium contents of the four top 7.6 cm layers were estimated to be 49.9, 46.5, 47.8 and 56.9 mg, prior to effluent irrigation (Table 5).

The amounts of sodium acetateextractable strontium remained between approximately 25 and 75 mg following treatment with the exceptions of the 7.6 to 15.2 cm layers of the C and 1F1D-1F3D treatments which contained 100.4 and 2.2 mg, respectively.

No strontium was removed in the grass clippings from any of the treatments.

Strontium found in the water samples was probably derived in part from the soil material in the columns with only part of the amount passing through the columns being from the applied effluent.

Although decreases of only 20.3, 3.6 and 16.4 percent and an increase of 3.2

percent were observed in the strontium content of the filtrate at the

243.8 cm depth of the

C, C-2F5D,

1F1D-1F3D and lFlD treatments, respectively, the average strontium concentrations of the filtrate were

estimated to be oniy 1.19 1.3, 0.8 and 1.0 mg/i.

The maximum weekly

85 concentrations for these respective treatments were estimated to be

3.1, 3.2, 1.1 and 2.3 mg/i.

Even though only small amounts of strontium were removed from the percolate by soil fiitration, the concentrations found in the filtrate were probably not undesirable for most water supplies.

No drinking water standard has been established for total strontium by the USPHS.

The radioactive isotope, strontium-90, was not evaluated separately in this study.

Strontium adsorption by soil was probably limited by the high amounts of sodium acetate-extractable strontium present in the soil at the inception of the study.

Also, the similarity between the physical and chemical properties of strontium and calcium ions may have lead to the replacement of strontium by calcium on many of the exchange sites in the calcareous soil.

Insofar as strontium removal from effluent is concerned, no apparent differences were observed between any of the four treatments employed in this study,

Even though most, or possibly all, of the strontium applied penetrated the entire length of the columns the quality of the resulting percolate was probably within acceptable limits for most water requirements.

Trace Element Removal at Grass Plot Site

Trace element investigations were conducted on the grass plot during the summer of 1967 but the site had been used for sewage effluent

An filtration studies for two years prior to the trace metal tests.

86 undetermined quantity and quality of effluent infiltrated into the soil during these experiments and undoubtedly account for differences between the grass plot results and those obtained from the lysimeter columns.

Sewage effluent was applied to the grass filtration strip at the rate of 870 cubic meters per day (225,000 gpd) during the two trace element tests.

The two and five week flooding tests were separated by a three week drying period.

The average concentrations of the trace metals detected in the filtrate at depths of 15.2, 30.5, 45.7 and 61.0 cm are listed in Table

7 and the maximum, minimum and average concentrations in the 9.1 and

15.2 meter well samples are listed in Table 8.

Unlike the lysimeter study, iron was present in the filtrate samples at all sampling depths.

The largest concentration, 0.053 mg/i, was detected at the 45.7 cm depth.

The average iron content in the upper saturated zone at 9.1 meters was 0.016 mg/i with only two of the five samples containing more than 0.008 mg/i of iron.

The average iron concentration at 15.2 meters was 0.006 mg/i.

Manganese concentrations in the upper 61.0 cm of soil at the grass plot were approximately ten fold higher than the averages obtained at comparable depths of soil in the continuously flooded columns.

The average manganese concentrations in the filtrate at

243.8 cm in the lysimeter coiumns was iess than 0.00i mg/i compared to

0.015 mg/i and 0.004 mg/i in the samples from the 9.i and 15.2 meter wells, respectively.

Neither hexavalent chromium nor nickel was detected in any of the samples from the grass plot.

88

Copper concentrations decreased with depth in the surface 61 cm of soil and reached a minimum value of 0.004 mg/l in the samples from the 9.1 meter well.

The average concentrations increased to

0.011 mg/i in the groundwater reservoir at 15.2 meters.

The concentrations were five to eight times greater than those obtained from the continuously flooded columns, between depths of 30.5 and 61.0 cm, but the filtrate at the 15.2 cm depth of the grass plot contained in excess of 50 times more copper.

The high concentrations of copper in the filtrate at 15.2 cm may indicate that the surface soil has become saturated with copper, during the two year grass filtration studies, and filtration is occurring at lower depths.

The average value of

0.009 mg/i at the bottom of the lysimeter columns is more comparable to those found for the deep percolate samples at the grass plot site.

Although zinc was detected in the filtrate samples obtained from the surface 61 cm of soil, and in four of five samples from the

15.2 meter well, within the groundwater aquifer, none was found in the upper saturated zone at 9.1 meters.

Zinc concentrations at 15.2 and

30.5 cm were nearly the same as those found in the continuously flooded lysimeter columns.

Lead movement was not as apparent under the grass plot as in the lysimeter columns, probably due to leaching during the preliminary experiments.

No lead was found in the filtrate samples below 15.2 cm and in only one of three samples obtained at the 15.2 cm depth.

Cadmium was observed in low concentrations in the three samples from the 15.2 meter well but in none of the shallow percolate samples.

89

This coincides with the lysimeter study in which no cadmium was found to be translocaecj through the soil.

Cobalt was detected in the filtrate at the 15.2, 30.5, 45.7

and

61.0 cm depths.

Because cobalt was not found in the effluent samples at the time of this test it appears that the concentrations observed may have originated from a previous period of effluent irrigation or from the soil material.

Detectable levels of cobalt were not present in the water after percolation through the soil strata above the upper zone of saturation at 9.1 meters,

Strontium, as in the lysimeter columns, was present in larger concentrations than any of the other elements tested,

For example, average concentrations of 2.172 and 1.772 mg/i were detected in the

9.1 and 15.2 meter well samples, respectively.

The results of the grass plot study indicate that the soil at the site became saturated with several elements after three years of intermittent flooding tests of various durations.

The elements most easily translocated in the grass plot study were copper, zinc, manganese and strontium.

Some iron was also observed in the filtrate at the grass plot whereas only an insignificant quantity was detected in the water from the lysimeter columns.

Lead and nickel may also have been translocated in the water at the grass plot site during the initial periods of irrigation, prior to trace metal sampling.

However, at the time of the study these metals either were not present in the soil or were in water insoluble forms.

In the lysimeter studies virtually all lead and nickel movement was detected during the initial six weeks of irrigation,

90 which indicates that water soluble forms were rapidly leached from the system.

The concentrations of the various trace elements in the upper few meters of the groundwater aquifer were within the recommended limits established by the USPHS (1962) for drinking water and would meet most industrial requirements.

Probably, with alternate wet-dry cycles even larger quantities of waste water could be renovated to a comparable quality as that obtained under the long term irrigation periods used in this study.

Nitrogen Compounds

The organic nitrogen content of the applied effluent was between 13.0 and 17.3 mg/i with a nitrate-N content of less than 0.4

mg/i.

The ammonium-N content of the effluent was between 12.1 and

12.8 mg/i.

Insignificant quantities of nitrite were detected in the samples.

The nitrate content of the percolate in the lysimeter columns was measured at the 61.0 and 243.8 cm depths at various times during the study.

For the first five weeks, the nitrate-N content of the water extracted from the 61.0 cm depth of the continuously flooded columns was less than 1.8 mg/i.

In the final six weeks of this test, the nitrate-N content increased to as much as 10.5 mg/i but all except two samples were within the range of 3.3 to 6.2 mg/i.

The nitrate-N content of the filtrate samples from the 243.8 cm depth ranged from

3.4 to 5.9 mg/i during most of the study with only slightly lower levels during the first few days of irrigation.

The low initial nitrate

91 content at the bottom of the soil columns was attributed to the anaerobic condition caused by the high infiltration rates and resultant saturation.

As the surface soil layer became clogged and an organic mat formed on the surface the infiltration rates decreased, allowing drainage of the coarser sand materials in the lower part of the columns.

The nitrate levels increased in the more aerobic environment.

The increased nitrate content of the filtrate samples extracted at the

61.0 cm depth during the latter weeks of the test also indicated somewhat better drainage and aeration of this layer with time.

An increased population of nitrifying organisms with time could also have contributed to the increased nitrate content.

The nitrate-N content of the percolate from both the 61.0 and

243.8 cm depths increased to more than 10 mg/l in the columns employing the 2F5D cycles.

Slightly lower nitrate-N levels were noted on each of the second of the two days of consecutive irrigation but values exceeding 10 mg/i were common.

The nitrate levels during the first day of irrigation were partially the result of leaching of nitrate formed by the oxidation of reduced nitrogen compounds during the drying period and during the early part of the irrigation phase before the soil oxygen was depleted.

Less nitrate was available for leaching on the second day of irrigation and the lower soil oxygen content, than during the first day, restricted the nitrification process, resulting in lower nitrate contents on the second day of irrigation.

The nitrate-N content of the individual filtrate samples extracted from the 1F1D columns during the first six weeks ranged from

0.9 to 11.3 mg/i, with an average value of 5.1 mg/i.

As drainage of

92 these columns became impeded during the final weeks, the nitrate-N content of the water decreased to 1.3 mg/i.

The three 1F1D columns placed on 1F3D cycles after the sixth week reflected the improved aeration by the increased nitrate content of the water at the 61.0 cm depth.

Nitrate-N values generally exceeded

11 mg/i.

The filtrate samples from the 61.0 cm depth of the lysimeter columns during the tenth week were analyzed for ammonium and organic nitrogen.

Extremely large quantities of organic nitrogen,

78.0 mg/i, in the filtrate samples from the continuously flooded columns can be attributed to the lack of oxidation in the anaerobic surface layer of these columns.

The maximum ammonium-N content, 1.7 mg/i, was detected in the samples from the 61.0 cm depth of the 2F5D treatment.

The organic-N content of this sample was 7.4 mg/i.

The large quantities of reduced forms of nitrogen in these columns reflect the large quantities of nitrogen in the applied effluent,

Possibly the time interval between application and sampling was insufficient to allow for the oxidation of ammonium and organic nitrogen.

Smaller amounts of ammonium and organic nitrogen were found in the filtrate from the 1F1D columns.

Less total nitrogen was applied than in the 2F5D and 1F3D treatments and the frequent periods of drainage and reaeration probably contributed to the oxidation of the reduced nitrogen compounds.

Under the 1F3D treatment, the ammonium-N content was less than

0.2 mg/i and the organic nitrogen level was 1.1 mg/i.

The lower quantities of reduced nitrogen, than in the 2F5D columns, probably

93 resulted from the application of smaller amounts of total nitrogen in each flooding period and the more frequent periods of reaeration.

The alternate-day-flooded columns contained the least amount of nitrate and total nitrogen of the four treatments.

A combination of frequent periods of reaeration and low application rates of total nitrogen probably allowed sufficient reaeration of the soil for nitrification of the reduced nitrogen compounds, yet maintained a partially anaerobic environment required for the denitrification process.

Longer reaeration periods allowed the oxidation of larger quantities of reduced nitrogen compounds but the increased application rates provided more total nitrogen than could be nitrified by the organisms during the drying periods,

The large and undesirable quantities of nitrate appearing in the filtrate were also the result of the reaeration of the soil which limited denitrification losses.

Even though the nitrate-N content of the filtrate from the continuously flooded columns was within the USPHS (1962) drinking water standards

(10 mg/l), the large quantities of organic nitrogen make this an undesirable treatment since much of this nitrogen could be converted to nitrate in the presence of oxygen.

The same nitrate relationships were detected at the grass plot.

The maximum nitrate-N contents, as high as 11 mg/i, were found on the first day of irrigation then decreased rapidly to less than 0.5

mg/i in the following week.

The levels remained low throughout the continuous flooding test but increased after a period of reaeration.

No nitrate-N was detected in the water samples from the 9.1 meter well and only 1.2 mg/i in the samples from the 15.2 meter weii, in the

groundwater aquifer.

The organic-N and ammonium-N contents of the

9.1 meter well samples were 1.4 and 0.6 mg/i, respectively.

Only

0.7 mg/l of organic nitrogen was detected in the groundwater samples.

94

The animonium-N content was 0.6 mg/i.

These results show that even under continuous flooding, for periods up to five weeks in length, low nitrate and total nitrogen levels can be achieved before recharge of groundwater occurs, provided a zone of aeration exists between the saturated surface layer and the water table.

An aerobic environment is required for nitrification of the reduced nitrogen compounds before the nitrogen is lost by the processes of immobilization and denitrification.

Under the narrow grass plot oxygen could enter from the sides of the profile.

The organic-N and ammonium-N contents of the effluent applied to the grass filter were approximately 15.0 and 12.5 mg/i, respectively.

The nitrate-N content was less than 0.4 mg/l.

Reductions of less than

20 percent for organic-N and less than seven percent for ammonium-N were accomplished by overland flow on the 304.8-meter-long grass filter.

Nitrate-N increased to as much as 1.8 mg/l in the surface outflow from the grass plot.

Similarly small total nitrogen reductions were obtained by grass filtration in preliminary tests during the previous year

(Wilson and Lehman, 1966).

Coliform Organisms

Total coliform counts were made on the filtrate samples from the lysimeter column study and on filtrate samples from the grass plot.

No

95 differentiation between fecal and soil coliform organisms was at temp ted.

After the second day of irrigation, the continuously flooded columns contained approximately 110 organisms per 100 ml of water at the 61.0 cm depth compared to 30 organisms per 100 ml in the 1F1D columns at the same depth.

By the end of the fourth day of continuous irrigation, the six columns contained from 270 to 700 organisms per

100 ml while the six 1F1D columns contained an average of 20 per 100 ml.

Much greater numbers of coliform organisms were noted in subsequent weeks but accurate counts were not attainable.

By the termination of the test, estimates of coliform organisms exceeded one million per 100 ml of filtrate from the 61.0 cm depth of the continuously flooded columns.

Some coliforms were detected in the percolate at the 243.8

cm depth as early as the fourth day of the test but the number was small compared to that at the 61.0 cm depth.

At the grass plot site, coliform organism samples extracted at

15.2, 30.5, 45.7 and 61.0 cm depths showed an increasing number of organisms with depth.

This increasing trend with depth was also observed at Whittier Narrows and reported by McMichael and McKee (1966).

They hypothesized that fecal organisms were filtered out near the soil surface but that non-fecal coliforms, such as Aerobacter aerogenes, flourished in the presence of organic materials.

The coliform counts were consistently less than 20,000 per 100 ml of water in the upper

61.0 cm of soil during the initial days of the test

By the end of the five week flooding test, the number of organisms were nearly as high at the surface as at the lower depths.

96

The fine silt and clay layer between 61 and 76 cm under the grass plot could have contributed to the increased number of organisms at the lower sampling depths through mechanical filtration of the organisms and accumulation of organic materials.

It was also observed that the deeper soil layers contained more moisture from previous irrigations, making them more favorable to soil organisms, than the surface layer which was dried by the hot summer days and the transpiring bermudagrass.

Numerous grass roots penetrated to the 30 cm depth but few roots were found below this level.

Of the treatments employed in this study, coliforin reductions were best accomplished by alternate periods of wetting and drying.

The number of fecal organisms was not determined but on the basis of previous work (Stone and Garber, 1951; Laverty et al., 1961; McMichael and McKee, 1966) it can be assumed that most of the organisms detected below the soil surface were of non-fecal origin.

Chemical Oxygen Demand

The chemical oxygen demand was used as a criterion for determining quality changes in sewage effluent during grass filtration.

filtration was evaluated at the grass plot site in two continuous

Grass irrigation periods of two and five weeks duration.

The effluent was detained on the plot for approximately six hours.

The COD of the applied sewage, in the first of the two floodings, ranged from 200 to

360 mg/i.

These quantities were reduced to 130 to 255 mg/l by grass filtration (30 to 40 percent).

In the second grass plot flooding, of five weeks duration, lower percentage reductions were attained although

97 the COD quality of the outflow was improved over the first test.

The

COD of the runoff ranged from 105 to 160 mg/i compared to 165 to 265 mg/i in the applied effluent.

These results indicate that COD reductions to desirable levels for industrial and municipal use, or for direct groundwater recharge, cannot be achieved by prolonged irrigation through grass filters of the dimensions and under the climatic

conditions

as at the grass plot site, even in the presence of biological slime layers on the grass.

Much of the COD in the outflow was attributed to the large algal concentrations in the water rather than to dissolved and suspended solids.

Some method of harvesting the algae would possibly reduce the COD to tolerable levels for some water requirements.

Recommended Treatment for Soil Filtration of Sewage Effluent

Of the four treatment methods employed in this study, the best results for quality improvement of sewage effluent were obtained with the 1F3D treatment.

Some results obtained in this treatment were less desirable than those accomplished in other treatments but the advantages exceed the disadvantages.

Many of the limitations of the 1F3D treatment could be minimized by making adjustments in the irrigation pattern.

Some form of intermittent irrigation would be preferable over continuous inundation in view of the limited reduction in total nitrogen and number of coliform organisms in anaerobic environments.

The major difficulties with the intermittent irrigation treatments used in this study, such as the application and translocation of large amounts of trace metals and nitrogen compounds in the large volumes of water

percolating through the soil in one day's time, might be overcome by limiting the volume of water applied in each irrigation period.

For

98 example, under continuous irrigation, infiltration rates equivalent to between six and eight centimeters per day produced only small amounts of metal movement in the soil system.

Even better results might be achieved if applications at this rate were made on alternate days or every third day.

These studies suggest that trace metal removal from a domestic sewage effluent can be accomplished by soil filtration, if a proper irrigation pattern is employed.

On the sandy desert soils used in this work, with a low clay content, applications of approximately ten centimeters every third day should produce a filtrate, ultimately mixing with groundwaters, with low concentrations of trace metals.

In addition, such a rate would maintain a soil environment which favors nitrate and total nitrogen reductions, fecal coliform organism filtration and elimination of dissolved organic constituents from the percolating water.

The resulting blend of native groundwater and recharged water should be of such quality that it could be used for most purposes with a minimum of additional treatment,

At this application rate, and assuming an average yearly evapotranSPiration loss of 20 percent, the volume of groundwater recharge would be 32,680 cubic meters per hectare per year (32.0 feet per year).

The available data is not sufficient to estimate the filtering life of the soil system for trace metal removal from sewage effluent.

If grass or some other form of vegetation is grown on the soil surface, some of the trace elements applied in the effluent will be

99 assimilated by the vegetation.

Complete removal of any of the elements does not appear to be feasible without restricting the applied quantities of effluent to levels that would make land disposal an uneconomical practice.

However, even the limited assimilation of metals by the plants is useful in prolonging the filtering capacity of the soil provided the grass clippings are removed from the soil surface.

The principal value of grass is probably in improving or maintaining soil structure to allow more rapid drainage and better aeration of the soil, which creates a more favorable environment for waste water reclamation.

S1Jtll4ARY AND CONCLUSIONS

Soil and grass filtration of a domestic sewage effluent for trace metal removal was studied by applying waste water to twelve

2.44-meter-long lysimeter columns and to a half-acre plot planted to common bermudagrass.

The soil material used in the columns was predominantly silt, sand and fine gravel.

Four different irrigation patterns were employed to determine the most effective irrigation cycle for the renovation of waste water and to estimate the quantities of groundwater recharge that could be achieved.

The four treatments, applied to three columns each, were (1) continuous irrigation for 11 weeks,

(2) continuous irrigation for six weeks followed by two days wet-five days dry cycles for five weeks, (3) alternate day irrigation for 11 weeks and (4) alternate day irrigation for six weeks followed by one day wet-three days dry cycles for five weeks.

Water samples were extracted from the soil columns at eight depths to establish the soil depth at which the various trace elements were removed from the filtrate by the processes of adsorption, absorption and biological assimilation.

Samples of the applied effluent, filtrate, grass and soil were analyzed for iron, manganese, chromium

(hexavalent), nickel, copper, zinc, lead, cadmium, cobalt and strontium by atomic absorption spectrophotometric techniques.

Filtrate samples were also collected at shallow depths at the half-acre grass plot, from an upper saturated zone at 9.1 meters depth

100

101 and from within a groundwater aquifer at 15.2 meters.

These samples were used for trace metal, coliform organism and nitrogen determinations.

Analyses for chemical oxygen demand, nitrates, ammonium and organic nitrogen were conducted on the inflow and outflow samples from the 3O4.8-meter-long strip to evaluate the effectiveness of grass filtration for quality improvement of waste water.

In the lysimeter column study, trace metal removal from the applied effluent was primarily effected at, or near, the soil surface for iron, manganese, nickel, copper, zinc, lead and cadmium.

Some translocatiori of copper, zinc and cobalt was observed at the deeper sampling points, especially when high infiltration rates allowed the application of large quantities of water.

Strontium was not effectively removed from the percolating water.

Larger amounts of iron, manganese and copper were detected in the filtrate from the grass plot than at comparable depths of the soil columns but the concentrations observed in the upper saturated zone and in the groundwater reservoir were not objectionable, insofar as USPHS drinking water standards are concerned.

The increased amount of trace metal in the filtrate probably was the result of effluent applied to the grass plot over a period of two years.

Less nickel and lead were detected in the grass plot filtrate than in the soil columns, probably the result of leaching in the two previous years.

Zinc, cadmium, cobalt and strontium were present in approximately the same amounts in both experiments.

No chromium was detected in either study.

Total nitrogen reductions were best accomplished when only small amounts of nitrogen were applied to the soil columns in each

102 irrigation period and when reaeration of the soil occurred during the drying period.

Low nitrate and total nitrogen contents were found in the groundwater beneath the grass plot as a result of a combination of aerobic microenvironments, which favored nitrification of reduced nitrogen compounds, and anaerobic microenvironments, which favored denitrification.

No total nitrogen reductions were accomplished by grass filtration and a slight increase in nitrates was observed.

Coliform organisms were present in large numbers in the filtrate samples extracted at the 61.0 cm depth of the soil columns but tha majority of organisms were probably of soil origin.

Fewer numbers were found in the deeper percolate samples under alternate day irrigation than in the continuously flooded columns, indicating the advantage, of an aerobic environment for coliform reduction.

The chemical oxygen demand of the inflow to the grass plot was reduced to values as low as 105 mg/l by grass filtration but further decreases would be unlikely.

A substantial amount of the COD remaining in the outflow was attributed to the high algal concentrations, produced by the favorable light intensities, rather than to dissolved and suspended solids that could be removed by additional filtration.

Of the four treatments employed in this study, the most favorable results were usually obtained with the one day wet-three

Modifications of this treatment cycle, for the days dry treatment.

conditions of this experiment, to approximately ten cm of effluent applied once every third day would probably maintain the advantages of

103 the one day wet-three days dry treatment.

In addition, some of the problems encountered with long periods of saturation and the application of large quantities of water, trace elements, nitrogen compounds, organic materials, fecal organisms and numerous other components of sewage effluent might be minimized.

The following conclusions were drawn from the results of this experiment:

It is possible to produce a water that will meet the trace element and nitrate requirements of drinking water, established by the

United States Public Health Service, by applying sewage effluent to soil at low rates and allowing sufficient time between floodings for reaeration and organic matter decomposition.

For the conditions of this test, a rate of ten cm every third day would allow approximately

32,680 cubic meters per hectare per year (32.0 feet per year) of groundwater recharge, for an undetermined duration.

By maintaining an aerobic environment in a sandy alkaline soil, iron, manganese, nickel and cadmium and all but small amounts of zinc and copper can be removed from the filtrate at, or near, the soil surface.

Manganese in the soil is mobilized when reducing conditions are produced by two or more days of continuous irrigation.

Anaerobic

Movement of zinc conditions also favor nickel and lead mobilization.

is greatest when large quantities of water percolate through the soil, causing anaerobic conditions.

Strontium is not effectively removed from sewage effluent by filtration through alkaline soils such as that employed in this study.

104

Total nitrogen and nitrate decreases are best accomplished when a combination of aerobic and anaerobic environments are maintained.

The necessary conditions can be provided by use of a narrow filtering basin or intermittent irrigation,

Common bermudagrass will assimilate iron, manganese, copper, zinc and cadmium from the soil but not in sufficient quantities to indefinitely sustain the trace element filtering capacity of the soil.

The primary value of the grass is in building or maintaining soil structure, thereby improving drainage and reaeration of the soil.

Grass filtration will not produce as high a quality of water as soil filtration, under the conditions employed in this study.

The high total nitrogen content and chemical oxygen demand limit the use of this water to agricultural irrigation unless additional treatments are employed.

The high COD values in the outflow from the grass filter are partially due to algal blooms which result from the nutrient content of the waste water and the favorable light intensities.

REFERENCES CITED

Alexander, M.

1961.

Introduction to soil microbiology.

and Sons, New York,

472 pp.

John Wiley

American Public Health Association.

1965.

examination of water and wastewater.

AWWA and WPCF.

769 pp.

Standard methods for the

12th edition.

APHA,

Beeson, K. C,, L. Gray andM. B. Adams.

1947.

Absorption of mineral elements by forage plants:

I. The phosphorus, cobalt, manganese and copper content of some grasses.

J. Amer.

Soc. Agron. 39: 356-362.

Bocko, J.

1965.

Soil as a medium for sewage purification.

Gleboznawcze 15: 497-548,

From:

13396h.

Roczniki

Chemical Abstracts 65 (4):

Bremner, J. N.

1965.

Total, inorganic and organic nitrogen.

In:

Methods of soil analysis.

Part 2.

Chemical and microbiological properties.

C, A. Black et al.

(editors).

Amer.

Soc. Agron., Mono. 9: 1149-1255.

and K. Shaw.

1958, Denitrification in soil: II. Factors affecting denitrification.

J. Agric. Sci.

51: 40-52.

Broadbent, F. E.

1951.

Denitrificatlon in some California soils.

Soil Sci. 72: 129-137.

and B. F. Stojanovic.

1952.

The effect of partial pressure of oxygen on some soil nitrogen transformations.

Soil Sci. Soc. Amer. Proc. 16: 359-363.

Day, A. D., and T. C. Tucker,

1959.

Production of small grains pasture forage using sewage effluent as a source of irrigation water and plant nutrients,

Agron. J. 51: 569-572.

and M. G. Vavich.

1962.

City sewage for irrigation and plant nutrients.

Crops and Soils 14 (8): 7-9.

Day, P. R.

1965.

Particle fractionation and particle-size analysis.

In:

Methods of soil analysis.

Part 1.

Physical and mineralogical properties.

C. A. Blacketal. (editors).

Amer. Soc. Agron., Mono. 9: 545-567.

105

106

Foster, II.

B., Jr.

P.

C. Ward and A. A.

removal by effluent spraying.

Prucha.

1955.

Nutrient

J. Sanit. Eng, Div., Amer.

Soc. Civil Engrs, 91 (6):

1-12.

Hill, A. C., S.

J. Toth and F.

E. Bear.

1953.

Jersey soils and forage plants

Cobalt status of New and factors affecting the cobalt content of plants.

Soil Sci. 76: 273-284.

Laverty, F. B,, R. Stone and L. A.

Meyerson,

1961.

effluent,

J. Sanit. Eng, Div., Amer.

Reclaiming 1-lyperion

Soc. Civil Engrs. 87 (6):

1-40.

McMichael, F.

C. and J. E. McKee,

Whittier Narrows.

1966.

Wastewater reclamation at

Calif. State Water Qual, Cont, Board

Publ.

No. 33.

100 pp.

Merrell, J. C., A. Katko and H.

E. Pintler.

1965.

tion project, Santee, California--summary

The Santee recreareport, 1962-64.

U, S. Public Health Service Publ. No. 999-WP-27.

69 pp.

Mitchell, R.

L.

the soil.

1964.

Trace elements in the soil,

In:

Chemistry of

2nd edition.

F.

E. Bear (editor).

Reinhold, New

York.

515 pp.

Ng, S.

K. and C. Bloomfield,

1961.

Mobilization of trace elements in waterlogged soils,

Chem. and md, (8): 252-253.

1962.

The effect of flooding and aeration on the mobility of certain trace elements in soils.

Plant and

Soil 16: 108-135.

Olson, R. V.

1965.

Iron,

In:

Methods of soil analysis.

Part 2.

Chemical and microbiological properties.

C. A. Black et al.

(editors).

Amer. Soc. Agron,, Mono. 9: 963-973.

Platte, J. A. andV. M. Marcy.

1965.

Atomic absorption spectrophotometry--a new tool for the water chemist.

Proc. Amer.

Power Conf. 27: 851-857.

Porges, R. and G. J. Hopkins.

1955.

Broad field disposal of beet sugar wastes.

Sewage and Industrial Wastes 27: 1160-1170.

Parizek, R. R., L. T. Kardos, W. E. Sopper, E. A. Myers, D. E. Davis,

M. A. Farrell and J. B. Nesbitt.

1967.

Waste water renovation and conservation.

The Penna. State Univ. Studies No. 23.

71 pp.

Price, W. J.

1967.

Atomic absorption analysis of waters and effluents.

Effluent and Water Treatment J.

7: 218-224.

107

Reeve, R.

C. and E. J. Doering, for salinity appraisal.

1965.

Sampling the soil solution

Soil Sd, 99: 339-344.

Riley, P.

1967.

Personal communication on trace element content of common bermudagrass,

Erly-Fat Livestock Feed Co.,

Tucson,

Ariz.

Robeck, G. G., T. W. Bendixen,

1964.

W, A. Schwartz and R. L.

Factors influencing the design and operation of soil systems for waste treatment.

Woodward.

J. Water Pollution Control

Fed.

36: 971-983.

Searle, S.

S.

1949.

purification.

The use of grass filtration areas for sewage

Paper presented to the 12th annual conference of Sewerage Engineers and Operators.

4 pp.

Starkey, R. L.

1955.

Relations of micronutrients to development of microorganisms.

Soil Sci. 79: 1-14.

Stone, R.

1953.

Land disposal of sewage and industrial

Sewage and Industrial Wastes 25: 406-418.

wastes,

J.

and W. Garber.

1951.

Sewage reclamation by spreading basin infiltration.

Proc. Amer. Soc. Civil Engrs. 77, Sep., No. 87.

Thomas, R. E., W. A. Schwartz and T. W. Bendixen.

1966.

Soil chemical changes and infiltration rate reduction under sewage spreading.

Proc. Soil Sd. Soc. Amer. 30: 641-646.

Thorne, D. W. and H. H. Wiebe,

1957.

Solubility and plant utilization of micronutrients,

Amer. Assn. Adv, Sci. Publ. No. 49: 49-69.

Truesdale, G. A., A. E. Birkbeck and D. Shaw.

1964.

A critical examination of some methods for further treatment of effluents from percolating filters.

J. and Proc. Inst. Sewage Purification 8: 3-23.

Ulrich. W. F. and N. Shifrin.

1965.

Atomic absorption spectrophotometry.

The analyzer.

Beckman Instruments, Inc.

6

(2): 10-14.

United States Public Health Service.

1962.

Drinking water standards.

U. S. Public Health Service Publ. No. 956.

61 pp.

Wheatland, A. B. and B. J. Borne.

1961.

Some changes in polluted water during percolation through soil.

Water and Waste

Treatment J. 8: 1-6.

Wilson. L.

G.

1967.

Sediment removal from flood water by grass filtration.

Trans. Amer. Soc. Agric. Engrs. 10: 35-37.

108

Wilson, L, G. and G.

S. Lehman.

1966.

Grass filtration of sewage effluent for quality improvement prior to artificial recharge.

Paper presented to the 1966 winter meeting of the Amer. Soc.

Agric. Engrs.

21 pp.

1967.

Reclaiming sewage effluent.

Progressive

Agric. in Ariz, 19 (4): 22-24.

Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement