WATER POLLUTION CONTROL RESEARCH SERIES
13020 DGX 08/71
ROLE OF ANIMAL WASTES IN AGRICULTURAL
LAND RUNOFF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
WATER POLLUTION CONTROL RESEARCH SERIES
the Water Pollution Control Research Series describes the
results and progress in the control and abatement of pollution
in our Nation*s waters. They provide a central source of
information on the research, development, and demonstration
activities in the Environmental Protection Agency, through
inhouse research and grants and contracts with Federal, State,
and local agencies, research institutions, and industrial
organizations.
Inquiries pertaining to Water Pollution Control Research Reports
should be directed to the Head, Publications Branch (Water),
Research Information Division, R&M, Environmental Protection
Agency, Washington, D.C. 20460.
-------
ROLE OF ANIMAL WASTES IN AGRICULTURAL LAND RUNOFF
by
Department of Biological and Agricultural Engineering
School of Agriculture and Life Sciences
North Carolina State University at Raleigh
for the
ENVIRONMENTAL PROTECTION AGENCY
Grant 13020 DGX
August, 1971
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $1.25
-------
EPA Review Notice
This report has been reviewed by the Environmental
Protection Agency and approved for publication. Approval
does not signify that the contents necessarily reflect
the views and policies of the Environmental Protection
Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendations for
use.
ii
-------
ABSTRACT
Twelve typical agricultural areas representing three types of
animal waste management techniques—lagooning, direct discharge into
streams and land spreading including pasture and drylot units—were
studied to determine the amounts of and factors governing stream
pollution from swine, dairy, beef, and poultry production operations.
More than 1500 stream and lagoon effluent samples were collected
with an automatic sampler developed for the study. The samples were
analyzed for bacteria, nutrients, and degradable organics. Hydrologi-
cal and waste management data were also collected.
Study results point to the superiority of land spreading for the
disposal of animal wastes. Good soil and water conservation practices
should be used to minimize the movement of wastes into streams. Higher
rates of runoff result in heavier pollution. The location of disposal
areas away from streams is important in controlling the amount of
entering wastes. Even when land disposal areas are poorly located, the
amount of pollution entering streams is usually low; and watershed
factors, such as surface culture and ease of erosion, are of primary
importance in governing the magnitude of pollution which reaches the
streams.
Effluents from swine waste lagoons were found to exceed raw
domestic sewage in strength and should not be discharged without
further treatment. Direct dumping of animal wastes into streams is
completely unacceptable and should be prohibited.
This report was submitted in fulfillment of grant 13020 DGX
between the Environmental Protection Agency and North Carolina
State University at Raleigh.
Key Words: Animal wastes, farm wastes, runoff, agricultural wastes,
water pollution, farm lagoons
iii
-------
TABLE OF CONTENTS
Page
CONCLUSIONS 1
RECOMMENDATIONS 3
INTRODUCTION 5
REVIEW OF LITERATURE 7
Farm Animal Situation 7
Swine 7
Cattle 8
Poultry 8
Properties of Animal Wastes 9
Swine 9
Cattle 10
Poultry 10
Related Studies 11
PROCEDURAL CONSIDERATIONS 15
Determination of Representative Animal Production Operations . 15
Site Data Collection Plan 16
Site Description and Management Practices l6
Swine Sites 16
Dairy Sites 23
Poultry Sites 27
Beef Site 27
Sample Collection Plan 31
Swine 31
Dairy 32
Poultry 32
Beef 32
Analyses of Samples ....... 33
Standard Analyses 33
Total Organic Carbon Analyses 3^
Nutrient Analyses 35
-------
TABLE OF CONTENTS (continued)
Page
EXPERIMENTAL DATA AND RESULTS ................... 39
Introduction .......................... 39
Experimental Data ............... • ....... 39
Simple Statistics and Regression Analyses ........... 40
Multiple Regression Analyses for Data from Swine Sites .... 43
Swine Waste Studies
Lagoon Studies ......................... 46
Direct Discharge Study .................... 60
Land Runoff Studies ...................... 68
Dairy Waste Study ........................ 82
Poultry Waste Study ....................... 83
Beef Waste Study ........................ 88
SUMMARY ..... ........ ................. 91
ACKNOWLEDGEMENTS ..... ........ . ........... 96
LIST OF REFERENCES ............ ........ .... 97
APPENDIX ............... . ............. 104
vi
-------
LIST OF FIGURES
Page
1 Map of Site D (Swine Study) 18
2 Rainfall and Amount of Swine Wastes Spread on Site E (Swine
Study) 20
3 Map of Site E (Swine Study) 21
4 Map of Site F (Swine Study) 22
5 Map of Site K (Swine Study) 24
6 Map of Site H (Dairy Study) 25
7 Map of Site J (Dairy Study) 26
8 Map of Site P (Poultry Study) 28
9 Map of Site X (Poultry Study) 29
10 Map of Site Z (Beef Study) 30
11 Representative Long-Term Lagoon Effluent BOD 41
12 Hydrological Data, Site A (Swine Study) 48
13 Bacteria Counts, Site A (Swine Study) 49
14 BOD5 and TOG, Site A (Swine Study) 50
15 Kjeldahl and Ammonia Nitrogen, Site A (Swine Study) 51
16 Phosphate, Site A (Swine Study) 52
17 Nitrite and Nitrate Nitrogen, Site A (Swine Study) 53
18 Solids, Site A (Swine Study) 54
19 Hydrological Data, Site D (Swine Study) 62
20 Fecal Coliform, Site D (Swine Study) 63
21 BOD5 and TOG, Site D (Swine Study) 64
22 Kjeldahl and Ammonia Nitrogen, Site D (Swine Study) 65
23 Phosphate, Site D (Swine Study) 66
24 Nitrate Nitrogen, Site D (Swine Study) 67
25 Hydrological Data, Site E (Swine Study) 70
vii
-------
LIST OF FIGURES (continued)
Page
26 Bacteria, Site E (Swine Study) 71
27 BOD5 and TOC, Site E (Swine Study) 72
28 Kjeldahl, Ammonia and Nitrate Nitrogen, Site E (Swine Study) . 73
29 Phosphate, Site E (Swine Study) 74
30 Circuitry for Automatic Sampler • 106
31 Advancing Mechanism for Automatic Sampler 107
32 Cut-away View of Automatic Sampler 108
viii
-------
LIST OF TABLES
Page
1 Correlation coefficients, r, for TOG with other pollution
indices .... ..... 43
2 Rates of effluent flows from the lagoons (swine study) .... 55
3 Bacteria in effluents from the lagoons (swine study) 56
4 BOD,, and TOG in effluents from the lagoons (swine study) ... 58
5 Nitrogen in effluents from the lagoons (swine study) 59
6 Phosphate in effluents from the lagoons (swine study) 60
7 Summary o'f the effect of discharging swine wastes directly
into stream D (swine study) . 69
8 Eates of streamflow from the land runoff sites (swine study) . 76
9 Bacteria in streamflow from the land runoff sites (swine
study) 77
10 BODc and TOG in streamflow from the land runoff sites (swine
study) 79
11 Nitrogen in streamflow from the land runoff sites (swine
study) 79
12 Phosphate in streamflow from the land runoff sites (swine
study) 80
13 Response of stream pollution to flow rate - March 18-25, 1969 . 82
14 Summary of surface water pollution from the dairy sites .... 84
15 Summary of surface water pollution from the poultry sites ... 87
16 Summary of surface water pollution from beef site . . 89
17 Correlation of TOG with other pollution indices for land
runoff studies . ........ 92
18 Average amounts of stream pollution from control watershed . . 94
19 Average amounts of wastes in land runoff from representative
animal growing operations 94
ix
-------
ABBREVIATIONS
Symbol Name
BOD biochemical oxygen demand
BOD- five-day BOD
COD chemical oxygen demand
FC fecal coliform
FS fecal streptococci
H number of unit hogs (100 Ib live weight)
M mean
NH_ ammonia nitrogen
N07 nitrite nitrogen
N0» nitrate nitrogen
No. number of values
OP orthophosphate
q flow rate
q total flow rate
q, base flow rate
q runoff flow rate,
-------
CONCLUSIONS
1. The natural pollution load on streams draining agricultural basins
free of farm animals can be appreciable during periods of rainfall
and runoff and should be taken into consideration in water quality
management.
2. Except for nitrate penetration into the groundwater at one site,
pollution indices for land drainage from waste spreading and con-
trol watersheds paralleled stream hydrographs with extended drag-
out on cessation of surface runoff. The rise in indices with run-
off was roughly proportional to increase in flow of stream over
base flow. Where nitrate had entered the groundwater, concentra-
tions in stream were inversely proportional to flow, peaking under
dry-weather conditions.
3. The extent of water pollution caused by farm animal production
units is more dependent on production and waste management practices
than on the volume of wastes involved.
4. The land provides a natural treatment system for animal wastes and
land spreading is a very effective means to prevent water pollution.
Even in cases where the disposal sites are poorly located or
managed or where pastured animals have access to streams, the
amount of pollutants (natural plus animal wastes) which reach
streams is a very small proportion (less than 10 percent) of the
potential from the animal wastes deposited in the watersheds.
Proper land spreading can reduce pollutants entering streams by
more than 99 percent. Criteria for this purpose are provided.
5. Differences in watershed characteristics such as slope, soil
permeability, surface culture, drainage pattern, degree of erosion,
and other factors are of great significance in determining the
quality of streams draining agricultural basins. This emphasizes
the importance of good soil and water conservation practices to
minimize the movement of wastes into streams.
6. Although estimating equations developed from this study with tem-
perature, number of animals and rate of land runoff as independent
variables and pollution parameters as dependent variables do not
have general applicability, predictive relationships held quite
well for many sets of data collected over short periods of time
showing promise that estimating equations to serve the needs of
water quality management can be developed with a more detailed
and longer term study, particularly if the equations include
effects of more hydrological variables.
7. The use of anaerobic lagoons as the sole means for treatment of
animal wastes is an unsatisfactory practice in areas where rain-
fall exceeds evaporation. Even when lagoons provide more capacity
-------
per animal than USDA and other recommended standards recommend,
effluents still exceed raw domestic sewage in strength. Although
the amount of surface discharge and resulting stream pollution
from lagoons can be lessened by reducing the amount of wash-water,
diverting runoff from surrounding areas, and locating lagoons to
prevent surface and subsurface inflow, at least intermittent sur-
face discharge is assured unless deep seepage is excessive.
8. The practice of dumping fresh animal wastes directly into streams
causes severe pollution. Swine and dairy production units are
the principal sources in North Carolina. Although the water quality
downstream from a discharge point is largely predictable from
characterization of fresh wastes, the quality varies erratically
with flow rate depending on the amount of solids carried by the
water. The large pollution load imposed on a stream by a direct
discharge operation overshadows the load from surrounding pastures
or other land disposal operations. The streams are generally more
polluted in the summer and pollution increases with surface runoff.
9. Simple regression analyses support the conclusion that total
organic carbon can be used as a rapid and reliable measurement of
pollution from animal wastes and for the estimation of other
pollution indices.
10. Antibiotics and toxic metals in animal feeds apparently interfere
with the BOD5 analysis of animal wastes at levels above 60 mg/1,
necessitating the concurrent use of TOG (or COD) for the estimation
of degradable organics and oxygen demand at BOD,- levels above 60 mg/1.
11. The state of the art of animal waste management for pollution con-
trol is primitive, indeed. Many questions remain with regard to
animal waste characterization, related water quality studies, and
the proper design of lagoons and other waste treatment facilities
used in conjunction with or independent of land spreading.
-------
RECOMMENDATIONS
1. Natural stream pollution should be taken into consideration in water
quality management. The extent of pollution caused by farm animals
should be distinguished from that coming from other pollution
sources that naturally occur in watersheds.
2. In view of the many desirable characteristics of land disposal
systems, adequate resources should be expended to improve their con-
ponents and to promote their widespread use. Pilot systems should
be established to resolve operational problems, improve system
design, and demonstrate effective management practices.
3. The direct discharge of animal wastes to water courses should be
prohibited.
4. The effluent from anaerobic lagoons should receive further treatment
or be disposed of by land spreading.
5. The total organic carbon test is recommended as a key index in
animal waste studies and surveys and for the estimation of other
pollution indices.
6. Additional research should be undertaken to:
a. provide additional factual data on the various forms and
extent of pollution from natural sources,
b. develop reliable standards for anaerobic and aerobic
lagoons with economical secondary treatment units
such as effluent irrigation and plant-soil filters for use
where circumstances preclude classical land spreading,
c. develop analytical and stream survey procedures for animal
waste,
d. determine the true extent of groundwater pollution asso-
ciated with animal wastes giving consideration to the
areal distribution of the pollution, and
e. develop generally applicable estimating equations that
include the essential variables for predicting the amount
of stream pollution arising from various animal production
and waste management units.
7. The following criteria are tentatively recommended for the guidance
of land disposal:
-------
a. apply wastes uniformly,
b. govern rate, time, and frequency of application for maxi-
mum nutrient utilization by plants,
c. select disposal areas with low erosion potentials,
d. do not apply waste on grassed waterways or other drainage
paths,
e. plow waste under on barren fields,
f. locate drylots away from streams and hillsides leading
directly to streams,
g. provide at least 100 feet of vegetated area between dry-
lots and streams or drainage paths, and
h. when stocking at a high rate, pasture animals away from
streams and drainage paths.
In essence, these factors are met by locating the production areas
according to hydrological dictates and by following standard soil
and water conservation practices.
-------
INTRODUCTION
General Background
This report presents findings of a study initiated to quantify
pollution loads in runoff from farm animal production areas and to
evaluate factors governing the timing, volume, and concentration of
waste discharges. The overall study is an attempt to determine whether
and under what conditions wastes from swine, dairy, beef, and poultry
production areas are a significant source of stream pollution and to
provide information that will aid in water quality management.
Agriculture is increasingly alleged to exert a detrimental influ-
ence on water quality (Wadleigh, 1968). Effluents from farm animal
production areas are particularly susceptible to generalizations that
mark them as a major contributor to surface water pollution. For
example, the potential significance of animal wastes as a pollutant is
often viewed in terms of human waste equivalents. The population equiva-
lent is about 16 for cattle, 2 for swine, and 1/7 for poultry (Environ-
mental Pollution Panel, 1956). For the nation as a whole, farm animals
produce ten times as much wastes as the human population. In North
Carolina the wastes from farm animals are equivalent to wastes from a
population of more than 15 million (Howells, 1969). While such generali-
zations may carry significant implications in water quality management,
they are hazardous if taken out of context. The extent to which animal
wastes are attenuated before reaching water courses has not been defined
by field tests or surveys. Knowledge concerning the actual pollution
imposed on streams from point and diffused sources of animal wastes is
necessary to properly assess the problem and formulate meaningful waste
control, treatment, and disposal requirements for surface water protection.
Eutrophication of lakes, reservoirs, and estuaries has led to specu-
lation as to the importance of effluents from animal production areas as
a source of nutrients, in particular nitrogen and phosphorus. Various
federal, state, and local authorities have advocated controlling the
phosphorus content of waters to a level low enough to prevent algae
growth. The threshold value has been reported to be from 10~5 mg/1 to
10"-*- mg/1 of phosphate (Levin, 1967). There is evidence that water con-
taining phosphorus above the threshold value will sustain increased
algae growth with added phosphate. In addition to its stimulation of
unwanted aquatic plant growth, nitrogen in its various oxidation states
is harmful to the health of aquatic and terrestrial animals when present
at sufficient levels. Fish kills occur when water contains high amounts
of ammonia (Rainwater and Thatcher, 1960). High nitrate levels in
drinking water may cause methemoglobinemia in humans (Bosch et^ al, 1950).
Abortions in cattle and sheep have occurred from ingestion of high
nitrate water (Wadleigh, 1968).
-------
Animal wastes are a concentrated source of bacteriological organisms
normally used to characterize water quality (Robbins e£ al, 1969b). The
dearth of investigations of relationships between water contamination
caused by animal wastes and disease transmission is somewhat of an anomaly.
Health problems associated with other means of contact with animal wastes
have been of concern to public health agencies for generations. A large
number of diseases of microbiological etiology are known to be transmitted
through animal wastes. One of the most important is salmonellosis
(Decker and Steele, 1966).
Animal wastes can deplete the dissolved oxygen in receiving streams
as the wastes undergo decomposition. Henderson (1962) concluded that
runoff from animal-growing areas can make municipal loads appear insig-
nificant on an oxygen-demand basis. Oxygen depletion in streams caused
by runoff from livestock feeding operations were considered to be the
primary cause of 15 of 27 fish kills in Kansas during 1964 (Smith, 1965).
Animal production is an important agricultural enterprise in the
Southeast from the standpoint of wastes produced and probable effects
on receiving streams. Animal waste management practices can be
categorized as land disposal (including pastures), lagooning, and direct
discharge to streams.
Objectives of the Investigation
The objectives of the investigation were to determine the extent
to which effluents from existing farm animal production operations con-
tribute to stream pollution and to evaluate the factors governing the
amount of pollution from these sources. Sub-objectives necessary for
the study were: (1) to develop equipment for the survey; (2) to develop
plans of sampling, sample handling, and sample analyses; and (3) to
determine the appropriate parameters for quantifying the pollution loads.
-------
REVIEW OF LITERATURE
The magnitude of the farm animal waste management problem was
pointed out in a report by the Secretary of Agriculture (1969) to the
President of the United States in which animal wastes in the United
States were stated to exceed wastes from any other segment of our
agricultural-industrial-commercial-domestic complex. The report esti-
mated that approximately 1.7 billion tons of animal wastes are produced
annually, about one-third of which is liquid. The federal budget for
fiscal year 1969 included over $2 million for research and development
to minimize pollution from animal wastes. Even larger sums are expected
to be available for expenditure each year for the next five years.
Farm Animal Situation
The value of livestock and poultry on the nation's farms and ranches
totaled $23.5 billion in January 1970. Meat animals accounted for $22.9
billion. The value of the cattle inventory was $20.2 billion while hogs
and pigs were valued at $2.2 billion. The chicken inventory was valued
at $581 million and turkeys at $36.8 million (NCDA, 1970). North Carolina
receipts from sales of livestock and livestock products in 1968 were
estimated at $488 million (NCDA, 1968a).
The number of farms growing animals has tended consistently down-
ward in the nation and in North Carolina (NCDA, 1968a). Animal produc-
tion is becoming more and more specialized as herds increase in size.
There has been a marked shift to confinement housing, coincident with an
increase in size of individual operations. Problems of waste management
such as stream pollution and economical disposal of wastes have occurred
because confinement housing has caused large numbers of animals to be
raised on relatively small areas and wastes to be concentrated in feeding
areas. Approximately 95 percent of livestock and poultry currently being
produced on North Carolina farms is marketed. Fifteen years ago one-
third of total production was consumed on farms where produced.
Swine
The swine industry produces about one-tenth of the total cash
receipts from farm products in the United States (NCDA, 1970). North
Carolina usually ranks twelfth in the nation in hog inventories and tenth
in pigs born. Production of hogs at the national level tends to follow
cycles over a comparatively short number of years (NCDA, 1968a). Produc-
tion in the Southeast and in North Carolina is generally consistent with
national cycles. Hogs on North Carolina farms have fluctuated between
peaks of slightly more than 1.5 million (record high of 1,675,000 in
1916) and valleys of slightly more than one million head (NCDA, 1968b).
-------
The December 1969 estimate of 56.7 million hogs and pigs on the
nation's farms was six percent less than the 60.6 million on hand a year
earlier. All hogs and pigs on North Carolina farms were inventoried at
1,612,000 head, up six percent from a year earlier (NCDA, 1970).
There were 3,691 farrowing houses, 601 nursery units, 270 sow
gestating units, 3,040 feeding floors, and 2,488 manure lagoons in North
Carolina during 1968 (Jones, 1968). About 40 percent of the hogs were
finished in concrete-floored houses. Regional distribution of swine
growing in North Carolina based on the sum of top hogs and feeder pigs
produced and sows kept for breeding, was: Coastal Plains - 75 percent;
Piedmont - 21 percent; and Mountains - four percent. There were 27,839
producers of hogs for sale, but only 7,000 with more than twelve farrow-
ing sows.
Cattle
Cattle and calves on the nation's farms and ranches were estimated
at 112 million as of January 1970. The North Carolina value was just
over one million head consisting of 318,000 milk animals and 763,000
beef animals. Milk animals in the Tar Heel State declined two percent
from the number of a year earlier.' An increase of six percent in beef
animals more than offset the decline in milk animals as has been the
case for the past decade. North Carolina ranks 35th in all cattle
(NCDA, 1970).
Regional distribution of cattle in North Carolina is:
Region Dairy Beef
Piedmont 55% 45%
Mountains 30% 26%
Coastal Plains 15% 29%
Poultry
At the beginning of 1970 chickens excluding commercial broilers on
the nation's farms were estimated at 432 million, up three percent from
a year earlier. Turkeys were estimated at 6.7 million, up one percent.
Chickens on North Carolina farms, excluding commercial broilers, totaled
23 million while turkeys totaled 769,000 head (NCDA, 1970). Commercial
broiler production in North Carolina was almost 263 million head in 1967.
The Tar Heel State ranks fourth in turkeys and chickens. Broilers are
second only to tobacco in cash receipts from farm marketings.
The distribution of poultry throughout the state is:
Region Layers Broilers Turkeys
Piedmont 63% 42% 25%
Mountains 10% 25% 5%
Coastal Plains 27% 33% 70%
-------
Turkeys are concentrated in the Southern Piedmont and Southern Coastal
Plains. Almost two-thirds of the broilers are from five counties repre-
senting the three regions of the state (NCDA, 1968a).
Properties of Animal Wastes
The properties of animal wastes can be classified as physical,
chemical, and biological. A great variability in properties is reported
in the literature. The variations may reflect different conditions of
production, management, climate and other factors. Also, today's con-
ditions and analytical techniques are radically changed from those of a
few years ago. Methods of collecting, handling, and analyzing the
wastes partially accounts for the wide differences in properties
reported. For example, animal wastes may be considered as only excreta
(solid and/or liquid) or may include litter or bedding, spilled feeds,
cleaning materials, chemicals used for pest control, cleaning water,
and waste carcasses. Analytical techniques for characterizing animal
wastes were adopted from those used for municipal wastes although animal
wastes are considerably more concentrated and of different composition.
Thus, analytical techniques have been suggested as a cause for the
differences (Taiganides and Hazen, 1966). Other causes for differences
in the reported properties of animal waste may be attributed to the
physiology (size, sex, breed, age, health, and activity) of the animal,
the feed ration (digestibility, protein and fiber content, and additives)
and the environment (temperature, humidity, light, antecedent moisture,
and time or age) (Robbins and Kriz, 1969).
Animal wastes contain large populations and varieties of bacteria
and viruses, some of which may be pathogenic to man and animals.
Gieldrich (1966) estimated that'the per capita production of the fecal
coliform indicator microorganism is:
hog 8,900x10 col/day
f
cow 5,400 x 10 col/day
chicken 240 x 10 col/day
Decker and Steele (1966) discussed the health aspects and vector control
associated with animal wastes. They indicated that many diseases organisms
originating in animal waste are readily waterborne.
Swine
Muehling (1969) concluded that the average BOD,, production is about
0.32 Ibs. per day per hog weighing 100 to 125 Ibs. He also reported that
the BODc for hog waste is approximately two-thirds the ultimate BOD, the
average chemical oxygen demand (COD) for a 100-lb. hog is about 1.4 Ibs.
per day, and the BOD5/COD ratio is usually about 0.3 for swine wastes.
Taiganides (1963) concluded that 60 percent to 70 percent of the
original amount of the ingredients in the feed eventually appear in the
fecal and urine excreta of the hogs. Muehling (1969) reported that the
-------
most generally accepted estimate of the amount of waste from a hog is
five percent to eight percent of its live weight per day of which 10 to
15 percent is dry matter. A 100-Ib, hog will produce approximately one
Ib. of dry solids per day. The wet manure per 1000 Ibs. live weight
hog averages 28,000 Ibs. per year; the major fertilizing elements
average: nitrogen (N), 175 Ibs; phosphorus ^05), 95 Ibs; and potassium
(K20), 128 Ibs. An average fertilizer value of hog manure is 0.7 percent
N, 1.4 percent P205» and 0.4 percent t^O. There is less urine than
solids, but the urine contains about 50 percent of the fertilizer
nutrients. Taiganides and Hazen (1966) reported that 1000 gal. of fresh
swine manure contains 47 Ibs of Ca, 6.6 Ibs. of Mg, 12 Ibs. of S, 2.3 Ibs.
of Fe, 0.5 Ibs. of Zn, 0.35 Ibs. of B, and 0.13 Ibs. of Cu.
Cattle
Wadleigh (1968) estimated that 400 cows will produce about 14 tons
of solid wastes and 4.5 tons of liquid wastes daily. Cattle wastes
were characterized by Witzel et_ al (1966). In addition to analyses of
the raw wastes, studies were made to evaluate the settleable properties
of the diluted wastes. Daily production values in pounds per 1000 Ib.
live animal weight were:
BOD5 COD TS N
Dairy Bull 0.76 4.19 4.21 .24
Dairy Cow 1.32 5.78 6.80 .37
Beef 1.02 3.26 3.26 .26
Physical properties of dairy animal wastes were described by Sobel (1966)
in terms of bulk density, partial size distribution, moisture content,
and solids composition.
Hemingway (1961) reported that 10 tons of cattle manure contains:
350 Ibs. ammonia sulphate, 120 Ibs. superphosphate, 100 Ibs. muriate of
potash, 68 Ibs. anhydrous Mg sulphate, 182 ppm Mn, 23.5 ppm B, 19.8 ppm
Cu, 1.7 ppm Co, and 2.3 ppm Mo. Herron and Erhart (1965) found that
manure from feedlots has 78 percent dry matter with 31 Ibs. N, 9 Ibs. P,
and 16 Ibs. K per ton. They concluded that the classical plant nutrient
analysis of 10-5-10 for cattle manures is not representative of manure
from western feedlots.
Poultry
Wadleigh (1968) estimated that one hundred thousand fowls produce
five tons of wastes per day. McHenry (1961) found that the weight of
manure voided by laying hens is approximately twice that of the feed
consumed. The weight at zero moisture content is about 1/2 the weight
of feed consumed. One thousand broilers produce 2.25 tons of dry drop-
pings during a ten-week growing period. One hundred turkeys produce two
tons during a 26-week growing period. Tower and Barr (1965) found that
the volume of manure from 1000 hens is 20 tons/yr.
10
-------
Hartman (1965) found that poultry manure dried to 12 percent
moisture contains 3.5 to 4 percent N. Phelps (1965) reported that fresh
poultry manure contains 1.8 percent N, 1.4 percent P, and 0.6 percent K.
Manure that is several years old contains 2.2 percent N, 1.1 percent P,
and 1.0 percent K. Other elements found in the manure include 2.2 per-
cent CaO« and some zinc.
Toth (1965) performed laboratory and field tests on dried poultry
excreta and bedding (bagasse). His results indicated that one ton was
equivalent to 1200 Ibs. of 5-10-10 fertilizer in terms of crop response.
Related Studies
Ludwig et al (1961) reported on urban and rural pollution in
Imperial Valley waters. Their data indicated that agricultural lands,
with and without farm animals, contribute significant amounts of bacterial
and BOD loading to surface waters. Henderson (1962) pointed to the
apparent neglect of the fate of animal excrement after deposition on
land surfaces with respect to their contribution to organic pollution
and bacteriological contamination of surface waters. He cited limited
data concerning the actual BOD contribution of agricultural land drain-
age to a stream. His data showed a correlation between rate of flow
and BOD^j. The lowest 6005 (0.6 ppm) was under drought condition when
the flow was 0.1 cfs per square mile. At a flow rate of 1.5 cfs per
square mile, the BOD5 reached its maximum observed value (3.6 ppm)
within the test series and maintained this value at the maximum observed
flow rate of 13 cfs per square mile.
Fish (1964) reviewed considerations for pollution prevention of
surface waters by animal wastes. To explain why this potential pollution
source has not appeared sooner as a major problem, he mentioned that
farm units are still very widespread and are often remote from rivers
and streams of consequence. He also cited four reasons why farm effluent
disposal could not be dealt with on the same basis as sewage and trade
effluent disposal: (1) farm effluents vary with time and between similar
farms; (2) operations on farms are quite alien to effluent purification
processes; (3) farm effluents generally cannot be purified to a high
standard of purity in conventional sewage plants; and (4) available
public sewage works are often incapable of handling much farm effluent.
Webster and Clayton (1966) concluded that organic pollution of
streams from farm animal wastes in terms of problems, processes, and
control methodologies has been neglected by water pollution, health, and
agricultural agencies.
Meiman and Kinkle (1966) studied the effect of impacted land use
(grazing and irrigation) on surface water quality. Bacteria indicators
gave a better indication of land use impact than did suspended sediment
or turbidity. All indicator organisms gave convincing evidence of
pollution on the grazed and irrigated watershed over and above that
occurring on a control watershed. Of the three indicator bacteria groups
used—fecal coliform (FC), total coliform (TC), and fecal streptococci
11
-------
(FS)—FC and TC were the most sensitive in detecting the grazing and
irrigation impact with FC being the best. The preliminary study indi-
cated that bacteria concentrations in the stream were related to over-
land flow, stream discharge, and season of the year. Storm character-
istics were very important in raising natural levels of organisms and
in magnifying organism differences in the natural versus impacted stream.
Ratios of FC/FS ranged from less than 1 to 4.5 on natural areas and less
than 1 to 44 on impacted areas. Maximum bacteria counts were obtained
during periods of lower flows and warmer water temperatures.
Weibel _et_ al (1966) included studies of pollution constituents in
runoff from five agricultural lands. Runoff from a cultivated field in
winter wheat that was free of farm animal wastes gave a BOD5 of 7 mg/1.
Nutrient levels from all five lands were many times the threshold levels
associated with algae problems. Bacterial indicator densities were high
in most farm land runoff samples. FS densities were greater than FC
densities from areas subjected to farm animal wastes. The authors took
this as an indication that the bacteria were predominantly from warm-
blooded animals other than humans.
Miner et^ al_ (1966a) evaluated the runoff-carried water pollutants
from feedlots. They encountered problems not applicable to domestic
waste sources because runoff is intermittent and is seldom confined to
a well-defined point. The runoff during and immediately after rainfall
was found to be high strength organic waste containing considerable
quantities of nitrogen. Concentrations of organic matter and nitrogen
increased with low rainfall intensities, warm weather, and moist lot
conditions. The amount of runoff from concrete lots was about twice
that from nonsurfaced lots. Runoff was heavily loaded with bacteria
normally used to evaluate water quality.
In reporting earlier work with Smith relative to water quality
measurements on water courses receiving feedlot runoff, Miner et al
(1966b) reported BOD5 values rising from 2 mg/1 to as high as 90 mg/1
following rainfall. High ammonia concentrations in runoff were also
noted. They found high bacterial counts and observed a decrease in the
FC/FS ratio when feedlot runoff was present.
Grub et al (1969) discussed in a general manner the factors affect-
ing the quality and quantity of runoff from feedlots. The factors given
were: (1) precipitation; (2) surface material; (3) land slope; (4) depth
of waste accumulation; (5) feedlot layout; and (6) ration composition.
Goleuke and Oswald (1966) indicated that the effluent from a strictly
anaerobic manure pond is not in a condition suitable for discharge into
surface waters since the BOD is excessively high (1100-3000 mg/1). Gen-
erally, they concluded that anaerobic ponds should not have any effluent.
Willrich (1966) and Curtis (1966) reported on the use of lagooning
for treatment of swine wastes. As with most other papers written on
animal waste disposal, primary emphasis was placed on freedom from odor
and other nuisance conditions as criteria for satisfactory operation.
Little or no mention was made of waste treatment to prevent water
pollution. Although Curtis was a public health engineer with the Nebraska
12
-------
State Department of Health, he cited only the following factors as
indicators of a successful lagoon operation: (1) a relatively odor-
free operation; (2) trouble-free water carriage of wastes away from
animals; and (3) an economic life in balance with initial investment.
In an investigation of swine waste disposal in Illinois in 1963,
Clark (1965) of the Illinois State Department of Public Health concluded
that "lagoons" were totally unsatisfactory, malfunctioning, open digesters.
The removal of solids by septic tanks and other means provided satisfactory
operation at a loading of 230 pounds of COD per acre per day. Available
information on swine waste strength was found to be unsatisfactory so
waste characterization studies were made with the following results:
BOD5, 8200 mg/1; COD, 67,200 mg/1; TS, 33,000 mg/1; VS, 26,000 mg/1;
volatile acids, 2,600 mg/1; coliform, 60 x 106/100 ml; enterococcus,
20 x 106/100 ml; pH, 6.5; NH^, 500 mg/1; N02, neg; N03, neg; N (Kjeldahl),
0.12 mg/1; and K^O, neg. 6005 tests were found to be unreliable because
of antibiotic effects. Chemical oxygen demand tests were consistent and
in general agreement with the 20-day BOD tests.
Loehr and Agnew (1967) and Loehr (1967) reviewed the literature on
cattle wastes and reported on anaerobic and aerobic disposal systems.
They concluded that the quantity and quality of wastes from beef feedlots
are such that a combination treatment system may be the most successful.
A combined anaerobic-aerobic lagoon system was thought to have signifi-
cant potential. The effluent from anaerobic lagoons was found to be
"potent" and required further treatment before discharge to a stream.
The successful disposal of liquid chicken manure by the plow-furrow
cover method was reported by Reed (1966). Additional studies to determine
the optimum capacity of the soil for manure disposal are underway.
In reporting on the requirements for microbial reduction of farm
animal wastes, Berry (1966) cited the low temperature of the manure
lagoon coupled with the high concentration of soluble materials as
limiting factors in its success. He expressed the opinion that manure
lagoons in the northern plains are at best primary settling basins with
little opportunity for a sequence of bacterial action. He saw little to
distinguish them from the domestic "seepage cesspools" used earlier in
the century.
Work by Irgens and Day (1966) on the aerobic treatment of swine
waste disclosed the following values for overflow from manure collection
pits (anaerobic lagoons): BOD, 1910 to 2890 mg/1; P04, 370 to 475 mg/1,
COD 4300 to 5200 mg/1 and N, 650 to 750 mg/1. The authors affected a
95 percent reduction in BOD with seven days aeration of effluent in the
laboratory. They have built oxidation ditches inside swine production
buildings, utilizing a continuous gutter under the self-cleaning slotted
floor. The effluent is then treated by secondary sedimentation before
disposal.
Trickling filters were found to be an effective means of reducing
the pollution qualities of dairy manure by Bridgham and Clayton (1966).
Their experiments suggested a final sedimentation tank volume of about
13
-------
3 3
114 ft /cow. A trickling filter system would require from 346 to 391 ft
of tanks per cow to produce an effluent BOD of 200 mg/1. (Note: The
strength is still 2/3 that of raw human sewage.)
No satisfactory assessment has yet been made of animal wastes as a
source of pollution of ground and surface waters. The approach to animal
waste treatment has been limited to fragmentary attempts at waste
characterization and nuisance prevention.
14
-------
PROCEDURAL CONSIDERATIONS
No precedent was available for a study to determine the extent to
which effluents from farm animal production areas contribute to stream
pollution and to develop the factors necessary to predict the amount of
pollution from these operations. Since the study was unique and differed
greatly from previous pollution investigations, little help in develop-
ing the research plan was found from a literature search. Thus, schemes
were proposed, tried, and modified as indicated by findings.
Representative animal production operations existing in the Pied-
mont Region of North Carolina were selected for study. The sites were
free of other sources of wastes except background (natural) pollution.
A control watershed free of domestic animal wastes was studied to define
the natural pollutional loading expected from the study sites in the
absence of animal production. Data detailing production and waste
handling facilities and practices were collected at each site. The
amount of effluents, or stream flow, from the sites were continuously
measured. Samples of the effluent from each site were collected and
analyzed to determine waste concentrations during warm and cool periods
and before and during periods of surface runoff.
Determination of Representative
Animal Production Operations
Nine North Carolina County Extension Offices (Wake, Chatham, Ran-
dolph, Johnston, Wilson, Edgecombe, Duplin, Harnett, and Cumberland)
were visited to obtain an understanding of the representative practices
of farm animal production and waste management. The major animal
producers throughout the counties were also visited.
The principal types of swine operations contributing to surface
water pollution were judged to be lagoons with surface outflow, drylot
operations, particularly those that allowed swine access to streams,
direct discharge of wastes from confinement growing areas to streams,
and drainage from areas used to receive wastes by spreading.
Representative dairy operations washed wastes from the milking
parlor into streams or drainage paths leading directly to streams. Some
solids were usually removed and spread on surrounding pastures.
The beef operations were primarily brood cow operations. Cattle
were generally pastured throughout the year and normally had direct
access to streams.
Two types of poultry operations, with respect to waste handling,
were found to be common in the counties visited. In one type, shavings
were added to the wastes (usually broiler operations but also a large
number of egg-production systems) while in the other no litter was
added. Both types of operations disposed of wastes by land spreading.
15
-------
Site Data Collection Plan
Data were collected from each production site to ascertain the
amount and strength of wastes produced and methods of waste disposal.
The following data were included: (1) description of animal-growing
facilities, including waste handling facilities and practices; (2)
number and size of animals; (3) type and composition of feeds, including
additives; (4) vegetation and cultural practices; (5) description of
drainage system from the production/disposal area to surface waters; and
(6) soil classification. Data for items 1-4 were supplied by the animal
operators on a monthly basis. When wash-water was used, the quantity
was estimated by measuring the rate of nozzle discharge and the period
of wash-down. Soil classifications were obtained from reports by the
Soil Conservation Service.
The hydrological data obtained for each area were: (1) rainfall
intensities and amounts, (2) amount of effluent or streamflow, and (3)
daily maximum and minimum air temperature. Water temperature measure-
ments were made when the sampling stations were visited. Rainfall was
measured at each site using recording rain gages. Flow measurements
were made with H-flumes and water-level recorders. These data were con-
tinuously recorded throughout the study periods. Temperature data were
from reports of the State College Weather Station, Raleigh, N. C.
Flow-through measurements for the lagoons were made during
February 1969. Flourescent dye (Rhodamine B) added to the influents of
the lagoons appeared in the effluents after one or two hours. The
monitoring was done with a 111-000 Turner Flourometer* The concentration
of dye in the effluent gradually increased and peaked after five hours
at site A and eight hours at the other sites. There was a very prolonged
drop-off in dye concentrations at all three lagoons.
Attempts made to determine the time of flow from land areas
receiving wastes to nearby streams were unsuccessful. Dye was applied to
the soil surface in the case of pasture operations and mixed with the
waste in the case of land spreading. The fluorometer reading of the
small amount of dye that may have reached the monitoring point in a
stream was confounded with an increase in background reading caused by an
increased sediment load carried by the stream during periods of surface
runoff.
Site Description and Management Practices
All twelve study areas were located in the Piedmont Region of
North Carolina near Raleigh. No change from the operator's normal
practices of either production or waste management was made for the study.
Swine Sites
Seven swine production sites were studied. Lagoons were used for
waste disposal at three of the sites, A, B, and C. Each lagoon had a
continuous surface discharge except during periods of prolonged absence
of rainfall. Swine sites, D, E, F, and K, were located in watersheds
16
-------
in which small streams originated. They included direct discharge of
wastes to a stream, waste disposal by land spreading, and an animal-
free control watershed.
The rations fed the swine were basically finely ground yellow corn
supplemented to contain about 15 percent protein, 0.7 percent calcium,
and 0.55 percent phosphorus. The feeds contained additives of mineralized
swine salt, vitamins, antibiotics, arsenicals, and antioxidants. A
particularly high rate (>1 Ib per ton) of copper sulphate was commonly
used. Cooked garbage was also fed at site A.
Site A. The lagoon at site A had been in use for six years and
served an average of 300 hogs from feeder-pig size (40 Ibs.) to market
size (225 Ibs.). It averaged five feet in depth and had a surface area
of approximately 50,000 ft . The lagoon received drainage from an addi-
tional 20,000 ft2 of surface area. The hogs were grown in confinement
on concrete and wastes were washed from the growing area each day. Wash
water averaged about 15 gal/day/hog. An estimated additional 15 gal/day/hog
was added to the lagoon from leaky valves and swine wastes. The topsoil
and subsoil in the area were tight clay.
Site B. Lagoon B was three years old and served an average of 100
hogs from feeder to market size. It was six feet in depth and about
20,000 ft2 in surface area. An additional 10,000 ft2 drained into the
lagoon. The hogs were grown in confinement on concrete and wastes were
washed from the growing area each day. The wash water averaged about
10 gal/day/hog. The subsoil in the area was heavy clay and the topsoil
was a sandy loam.
2
Site C. Lagoon C was two years old, eight feet deep, and 20,000 ft
in surface area. It drained an1 additional 5,000 ft2. The average hog
population previous to and during the study was 90. The hogs were grown
in confinement on concrete. Wash water was about 10 gal/day/hog. During
the cooler months, wastes were retained in a trench under the hog house
for periods as long as two weeks before being dumped into the lagoon.
The lagoon was located in a tight clay soil.
SiteD. At site D, wastes from a farrowing house, pens holding
feeder pigs, and runoff from lots used to pasture brood hogs entered the
stream that originated in the basin. The house and pens had concrete
floors. Water for drinking purposes flowed continuously through the
house and pens from a farm pond located at the head of the stream. The
wastes produced in the confined growing areas were washed directly into
the stream, normally every two days. The lots were about 10 acres in
size and were located on the steeper part of the watershed. Figure 1 is
a map of site D including soil types, slopes, and sampling station loca-
tion. About 70 percent of the area was void of vegetation. Small
gullies were predominant throughout the lots. The total drainage area
above the sampling point was about 40 acres. Corn and tobacco were grown
on 20 acres of the better land, and the remaining area was in woodland
and farm roads.
17
-------
NOTE:
Shaded area
is area of
drylots. X
is dumping
site.
Station D
SOIL CLASSIFICATION LEGEND
365B1 - Norfolk sandy loam, 2-6% slope, slightly eroded
365C2 - Norfolk sandy loam, 6-10% slope, moderately eroded
370A1 - Norfolk loamy sand, 0-2% slope, slightly eroded
370B1 - Norfolk loamy sand, 2-6% slope, slightly eroded
370C3 - Norfolk loamy sand, 6-10% slope, severely eroded
410B1 - Goldsboro sandy loam, 2-6% slope, slightly eroded
810 - Mixed local alluvial land, poorly drained, nearly level
Figure 1. Map of site D
18
-------
The production area had been used two years prior to beginning the
study. However, about eight months before the start of the study all
swine had been removed because of a disease problem. New brood stock
were obtained shortly thereafter, but the farrowing house and holding
pens were not used until farrowing began at the onset of the study. The
brood sow population averaged 150 on the lots and 50 in the farrowing
house during the study. After the first month of study an average of
450 feeder-size pigs were housed in the holding pens.
Site E. A small stream originated at the head of a gully at site E.
About five acres of the 35-acre drainage basin above sampling station E
were used to spread the wastes from swine growing in houses with partially
slotted concrete floors. The remainder of the watershed was in grass
(10 acres), small grains (5 acres), corn (10 acres), and orchards (5
acres). No farm animal wastes had been deposited in the study area for
six months before the study began. The only wastes on the watershed
prior to this was droppings from beef grazing the disposal area. A
liquid manure spreader was used to pump the wastes from the 4-ft. by
4-ft. ditches under the houses and to spread it on the grassed disposal
area. A minimum amount of water was used in the houses and the wastes
were removed from the pits only as they became filled. Normally, a
portion of the manure was removed every two weeks.
When spreading began, about 200 hogs averaging 100 Ibs. in size were
being grown to market size. After they were removed, another group of
about 500 were farrowed and finished out in the houses. The cumulative
rainfall and wastes spread between December 1, 1968, and June 18, 1969,
are represented in Figure 2. During this period the waste application
rate averaged about 4,000 gal/acre/mo. Three acres of lots for brood
sows were established in the drainage basin during the study period.
Eighty-six sows and 100 gilts were moved to the lots on February 20,
1969, and this number remained on the lots throughout the remainder of
the study period. The lots were well grassed initially, but had begun
to show some bare spots by June 1969. Figure 3 is a map of site E
including soil types, slopes, and sampling station location.
Site F. A small stream originated in the 75-acre basin above
site F, the control site. Site F was adjacent to site E and was devoid
of domestic animal wastes. No farm animal wastes had been spread or
deposited on the area for six months prior to the study period. The
basin consisted of 25 acres of grassed pasture, 20 acres of wood pasture,
20 acres of corn land, and 10 acres of orchard. About two acres was
swampy. Figure 4 is a map of site F including soil types, slopes, and
sampling station location.
Site K. A stream originated in the basin at site K. At the begin-
ning of the study, the site was subjected to domestic animal wastes from
three acres of drylot pasture holding 150 brood sows that had access to
the stream. The pasture had been in use for six months before sampling
began and was woodland with little surface cover other than large trees.
The pasture area was steep and had small gullies throughout. Just prior
to beginning the study, three acres of additional lots were established
in the basin. This area was well covered in grass and was gently sloping.
19
-------
24 r-
20
16
CO
0)
43
O
a
-S 8
cfl
O Rain
— Was te
12
10
ctf
O
c.
6 a)
4-1
CO
Dec
Jan Feb Mar Apr May Jun
Figure 2. Rainfall and amount of swine wastes spread on site E
20
-------
NOTE: Shaded area is
area of waste
spreading.
Hatched area is
area of drylots.
Station E,
SOIL CLASSIFICATION LEGEND
31B2 - Cecil sandy loam, 2-6% slope, moderately eroded
32B2 - Cecil fine sandy loam, 2-6% slope, moderately eroded
32C2 - Cecil fine sandy loam, 6-10% slope, moderately eroded
33C2 - Cecil gravely fine sandy loam, 6-10% slope, moderately eroded
36C2 - Applying sandy loam, 6—10% slope, moderately eroded
36D2 - Applying sandy loam, 10-15% slope, moderately eroded
37B2 - Applying fine sandy loam, 2-6% slope, moderately eroded
37C2 - Applying fine sandy loam, 6-10% slope, moderately eroded
40 - Colfax sandy loam
Figure 3. Map of site E
21
-------
Station F
SOIL CLASSIFICATION LEGEND
32B2 - Cecil fine sandy loam, 2-6% slope, moderately eroded
33C2 - Cecil gravely fine sandy loam, 6-10% slope, moderately eroded
36B1 - Applying sandy loam, 2-6% slope, slightly eroded
36C2 - Applying sandy loam, 6-10% slope, moderately eroded
36E2 - Applying sandy loam, 15-25% slope, moderately eroded
37B2 - Applying fine sandy loam, 2-6% slope, moderately eroded
40 - Colfax sandy loam
Figure 4. Map of site F
22
-------
The total hog population in fche basin reached 250 near the end of the
study. The remaining drainage area included about 20 acres of tobacco
and other row crops, 5 acres of unused grassland, and 20 acres of wood-
land. Figure 5 is a map of site K including soil types, slopes, and
sampling station location.
Dairy _S_i_tes_
Two similar dairy sites were studied. Both operations washed liquid
wastes from the milking parlor into drainage paths leading to small
streams which originated in the basins. Solids removed from the parlors
were spread on surrounding pastures. These drained into the streams
receiving wastes directly from the parlors.
Site H. The area at site H had been converted from row crops to a
dairy operation six months before the study began. The number of dairy
animals contributing wastes to site H was as follows:
Month Milking Dry Heifer > 1 yr Heifer < 1 yr
April 61 3 4 12
May 63 1 3 26
June 64 1 12 16
July 50 14 22 16
August 44 22 27 9
September 58 16 14 8
October 58 20 12 0
November 45 18 11 0
December 54 12 8 0
January 60 8 6 0
During April-August 1969, 8 tons/mo of 16 percent dairy ration and
40 tons/mo of corn silage were fed to the milking cows. During September-
January, 15 tons/mo and 60 tons/mo, respectively, were fed. Hay con-
sumption averaged 20 tons/mo throughout the study period.
An automatic scraper was used in the dairy parlor to remove solids
to a manure spreader.
The 70-acre watershed above station H consisted of 55 acres of
pastures and 15 acres of corn land. Figure 6 is a map of site H includ-
ing soil types, slopes, and sampling station location.
Site J. The area at site J had been used as a dairy for more than
10 years. An average of 110 milking cows, 15 dry cows, and 36 heifers
were kept in the 15-acre watershed. All but two acres (domestic dwell-
ings and shops) were pasture and barnyard. Figure 7 is a map of site J
including soil types, slopes, and sample station location.
Feed composition averaged 20 tons/mo of 3/4 part ground corn and
1/4 part 44 percent soybean, 360 tons/mo of sorgham sileage, 200 tons/mo
of corn sileage, and 10 Ib/day/cow of orchard and/or alfalfa hay.
23
-------
NOTE: Shaded area
is area of
drylots.
Station K
SOIL CLASSIFICATION LEGEND
365B1 - Norfolk sandy loam, 2-6% slope, slightly eroded
366B2 - Norfolk fine sandy loam, 2-6% slope, moderately eroded
370B1 - Norfolk loamy sand, 2-6% slope, slightly eroded
392A1 - Fuquay (Wagram) loamy sand, 0-2% slope, slightly eroded
49 - Mixed local alluvial land, moderately well drained
715B1 - Lakeland loamy sand, shallow phase (30-42 in.)» 2-6% slope,
slightly eroded
810 - Mixed local alluvial land, poorly drained, nearly level
Figure 5. Map of site K
24
-------
NOTE: Shaded areas
is pasture.
'airy parlor
Station H
SOIL CLASSIFICATION LEGEND
(See Figures 1, 3, and 4.)
Figure 6. Map of site H
25
-------
NOTE :
Shaded area
is pasture.
Station J
SOIL CLASSIFICATION LEGEND
31B6 - Cecil sandy loam, 2-6% slope, buildings
31E2 - Cecil sandy loam, 15-25% slope, moderately eroded
63E2 - Wilkes loam, 15-25% slope, moderately eroded
See Figures 3, 4, and 5 for other descriptions.
Figure 7. Map of site J
26
-------
An exceptionally poor job was done of removing solids from the
milking parlor before wash-downs, and a larger portion of the wastes were
washed directly into the receiving stream than from site H.
Poultry Sites
Two areas receiving wastes from egg production operations were
studied. The feed used was a standard ration in both operations, i^,e_.,
Corn 1200 Ib
Soybean 400 Ib
Meat scrap 60 Ib
Fish meal 50 Ib
Alfalfa 60 Ib
Limestone 120 Ib
Phosphate 40 Ib
Premix vitamins 5 Ib
Salt 8 Ib
Corn fermentation .... 50 Ib
About one-third of the feed had pro-strep, penicillin/streptomycin. Feed
consumption averaged 110 Ib/yr/hen.
Site P. Shavings were mixed with poultry wastes spread on site P.
Wastes were applied to the 65-acre watershed three times per year from
three houses containing 3500 laying hens each. Each house was cleaned
once per year and the wastes were applied to about five acres. The water-
shed had been used for waste disposal for more than ten years.
During the study, wastes were spread on the watershed twice. The
first spreading was during the first week of October, 1969. The second
was on January 13-14, 1970. The rate of.spreading was about four tons
per acre and spreading was on row and cover cropped fields, away from
meadow strips and other well-defined drainage areas. Figure 8 is a map
of site P showing soil types, slopes, and sampling station location.
Site X. Figure 9 is a map of site X showing soil types, slopes, and
sampling station location. The station was located in a meadow strip
draining the surrounding fields which had been turned during the fall of
1969.
Twenty-two tons of fresh poultry waste, not containing litter, were
spread on four acres of the five-acre watershed above station X on
February 10, 1970. The waste was spread on the plowed fields and in the
meadow strips. It weighed one ton/yd^.
Beef Site
Site Z was studied to determine the amount of pollution reaching
streams from beef pastures. Figure 10 is a map of the watershed above
station Z and shows the soil types and slopes. Fifteen acres of the
25—acre watershed were in pasture. The remaining area was woodland and
idle land. A small stream originated at the head of a gully in the
pasture.
27
-------
NOTE: Shaded area
is area of
waste
spreading.
iry houses
Station P
SOIL CLASSIFICATION LEGEND
11 - Local alluvial land, sandy
See Figures 1 and 3 for other descriptions.
Figure 8. Map of site P
28
-------
NOTE: Shaded area is area
of waste spreading.
Station X
SOIL CLASSIFICATION LEGEND
31B2 - Cecil sandy loam, 2-6% slope, moderately eroded
31C2 - Cecil sandy loam, 6-10% slope, moderately eroded
38B2 - Appling gravelly sandy loam, 2-6% slope, moderately eroded
Figure 9. Map of site X
29
-------
NOTE: Shaded area is
pasture area.
Station Z
SOIL CLASSIFICATION LEGEND
43 - Chewcla fine sandy loam, 0-6% slope
See Figure 3 for other description.
Figure 10. Map of site 2
30
-------
Twenty-three beef cattle averaging 500 Ib live weight were placed
in the pasture during October 1969 two months before sampling began.
The area was void of farm animal wastes during the proceeding eight
months. During January 1970 an additional 26 beef animals that averaged
600 Ibs. each were placed in the pasture until the end of the study,
May 1970. Most of the animals were heifers and were fed hay, silage, and
some ground corn and soybeans to sustain them through the winter.
Twelve samples were also collected from the stream at site F, the
control watershed, during a runoff period in late August 1970 two weeks
after 34 beef animals were placed in the pasture. These data are not
reported since the study period was short and waste accumulation small.
No change was noted in the quality of the water due to this small number
of animals per unit area of the watershed.
Sample Collection Plan
More than 1500 samples of the effluent from the twelve study areas
were collected in iced containers by automatic samplers described in
Appendix A. The samplers were started manually for the lagoon studies
and automatically when the water level rose to a predetermined level
for the stream studies. Once sampling was started, samples were collected
at selected time intervals. A common sample interval was two hours.
Samples were removed from the sample containers as soon as feasible
(within 24 hours) after collection and returned to the laboratory for
analysis. Samples collected manually were delivered to the laboratory
immediately. A number of samples were duplicates; i.e., two samples
collected at the same time to make procedural checks. Occasional samples
were composited before analyses.
Swine
Swine waste management practices involved include the three basic
means used to handle animal wastes in the Piedmont Region of North
Carolina; i.e., land disposal, lagooning, and direct discharge into
streams. Since installation of equipment was simpler at the lagoon
sites and sample collection was less dependent upon rainfall, sampling
of the lagoon effluents was started first in October 1968. This per-
mitted an early check-out of equipment and analytical techniques as well
as providing the required data from the sites. Sampling of lagoons B
and C was discontinued after the spring of 1969. The sampling at site A
was continued through the summer of 1969. Most lagoon samples were col-
lected during October, November, February, and March. About 50 percent
were collected on an hourly basis. Sampling during other periods was
generally on a weekly or bi-weekly basis. The numbers of samples taken
from the outflow from the lagoons and analyzed were 179 at site A, 90 at
site B, and 107 at site C. Fourteen grab samples of the inflow to the
lagoons were also collected.
Sampling at sites E and F began in December 1968 and was continued
until July. About one-fourth of the samples from these sites were
collected during periods of no surface runoff. These were used to evaluate
changes in water quality of the base flows as well as changes in water
31
-------
quality caused by surface drainage. The number of samples collected at
site E was 176 and at site F 148. Eighteen samples of the wastes being
spread on site E were also collected.
Sampling at station D began in March 1969 and was continued through
June. About one-half of the 117 samples were collected during periods
of no surface runoff. The samples were used to evaluate the load on the
stream caused by direct discharge of wastes from the farrowing house and
holding pens and to provide a base for establishing the contribution from
the drylots during periods of surface runoff.
At station K, 100 samples were collected from April to August 1969.
About one-fourth of the samples were collected during periods of no
surface runoff to ascertain the base flow quality, to measure contributions
from swine having access to the streams, and to aid in evaluating the
loadings during periods of surface runoff.
Sampling at site H began in June 1969 and continued to January 1970
while the study at site J covered the period from October 1969 to April
1970. About one-fourth of the samples from each of these sites were
collected during periods of no surface runoff. These provided a means to
evaluate pollution caused by washings from the milking parlors and to
provide a base for establishing the contribution from the pastures during
periods of surface runoff. The number of samples collected from site H
was 126 and that from site J 102.
Poultry
At site P, 108 samples were collected between June 1969 and April
1970. About one-fourth of these were collected during base-flow con-
ditions and provided a basis for evaluating the changes in water quality
of the base flows with time as well as changes in the stream quality
caused by surface runoff.
Sampling at site X began in February 1970 before wastes were spread
and continued through April 1970. The predisposal samples were collected
to determine the natural pollution load. Samples were collected during
the first runoff period shortly after the spreading operation to ascertain
the maximum contribution expected from land disposal of poultry wastes.
Subsequent samples were collected in April to determine the decay of
pollution with time after spreading.
Beef
At site Z, 78 samples were collected from November 1969 to April
1970. About one-fourth of these were collected during periods of no
surface runoff to ascertain the base flow quality, to measure contributions
from beef animals having access to the stream, and to aid in evaluating
the loadings during periods of surface runoff.
32
-------
Analyses of Samples
Generally, the analyses performed on the samples were according
to procedures described in Standard Methods (APHA, 1965). Analyses
performed on a particular sample were in keeping with the objectives of
the study and the work load in the laboratory. A full-time technician
performed all tests except nutrient analyses. Analyses for the phos-
phorus and nitrogen series were performed by the North Carolina Depart-
ment of Water and Air Resources using an Auto-Analyzer.
Analyses included the following:
1. Total coliform (1C), colonies/100 ml.
2. Fecal coliform (FC), colonies/100 ml.
3. Fecal streptococci (FS), colonies/100 ml.
4. Biochemical oxygen demand (BOD), mg/1.
5. Chemical oxygen demand (COD), mg/1.
6. Total solids (TS), mg/1.
7. Volatile solids (VS), mg/1.
8. Total organic carbon (TOG), mg/1.
9. Kjeldahl nitrogen (TN), mg/1.
10. Ammonia nitrogen (NE^), mg/1.
11. Nitrite nitrogen (N02), mg/1.
12. Nitrate nitrogen (N03), mg/1.
13. Total phosphate (TP), mg/1.
14. Orthophosphate (OP), mg/1.
Specific conductivity and pH measurements were also made on several
samples. The following pollution indices were determined to be most
suitable for delineating the quality of the samples and were the principal
analyses performed during heavy laboratory workload periods: (1) FCj
(2) BOD5; (3) TOG; (4) TN; and (5) TP.
Standard Analyses
Density of bacteria indicator organisms was measured by the mem-
brane filter procedure (Taylor et_ al_, 1955). TC counts were made using
M-Endo medium and incubation at 35°C for 24 hr (Kabler and Clark, 1960).
FC counts involved incubation at 44°C for 24 hr and a M-FC medium
(Geldreich et_ al, 1965). KF medium was used for FS measurements with
incubation at 35°C for 48 hr (Kenner £lt a.1, 1961). Although primary
reliance was placed on the FC to indicate bacterial contamination, a
large number of samples from the swine sites were analyzed for TC
because of the wide acceptance of this test. Because the FC/FS ratio
had been suggested as an indicator of waste origin (Geldrich, 1966), a
few FS measurements were made.
Because several investigators, e.g. Clark (1965), had reported
difficulties in using the BOD parameter to characterize animal wastes,
plans were made to place primary reliance on TOC values once BOD/TOC
relationships were determined. However, poor BOD results for the lagoon
samples made establishment of any BOD/TOC relationships questionable and
the decision was made to continue the BOD throughout the study (Robbins
et al, 1969a).
33
-------
Most BOD tests were standard, five-day, 20°C, unseeded, determina-
tions. Seeding and long-term tests were performed on a few samples.
A Weston and Stack dissolved oxygen meter was used to measure the dis-
solved oxygen in most BOD tests. This was calibrated and checked with
the Winkler method. BOD results were found to be dependent on the
amount of sample used, particularly for the more polluted samples.
Usually, the results showed higher BOD values in greater dilutions.
Although these BOD variations were discovered early in the study, a
thorough review of procedures failed to account for them in terms of
procedural errors. The variations were attributed to the characteristics
of the test and of the wastes. Multiple tests and additional dilutions
(above the standard three) were performed on most samples collected
after November, 1968. Some samples were analyzed for BOD immediately
after collection and after one to six months of frozen storage.
Determinations of TS and VS were made on more than 100 lagoon
samples. Because of low solids contents, solids tests on stream samples
proved too inaccurate to be of use in assessing the amounts of wastes
reaching the streams and were not generally performed on these samples.
Although not included in the early scheme of data collection, COD
analyses were performed on about fifty of the lagoon effluent samples.
Since good correlations between COD and TOG existed for samples of
influents and effluents of the lagoons, the COD test was. performed only
as the heavy laboratory workload permitted. Often the COD values of the
stream samples were too low to allow the test to be used as a tool for
monitoring the contribution of wastes to the streams. However, the COD
test was quite applicable to the more polluted stream samples and was
used rather extensively during the dairy, beef, and poultry studies.
Total Organic Carbon Analyses
Two basic avenues toward estimating the oxygen demand of animal
wastes are: (a) to determine the amount of oxygen or equivalent oxidant
required to react with the organic materials, and (b) to measure the
amounts of oxygen-demanding materials in the wastes. BOD and COD tests
employ the oxygen-based approach. While both have wide acceptance, they
have major inherent limitations. The TOC analysis offers many potential
advantages for the measurement of organic waste substances. It elimi-
nates many of the variables that confound oxygen-based tests, particularly
the BOD tests, and produces more reliable and reproducible results.
There are potential savings in analytical time and expense when the TOC
test is used. A distinct advantage is the reduced time lag between
sample collection and analysis. It also provides a carbon balance
procedure that is useful in characterizing and following the decomposition
of waste organic substances with time and treatment. The major limita-
tion of the TOC analysis is the fact that the results are unfamiliar to
investigators and a correlation of carbon-based results and oxygen-based
results is necessary if it is to have direct use for waste characteriza-
tion and water quality monitoring.
Until recently, the measurement of low concentrations of carbon in
aqueous solutions was difficult and time consuming. For example, the
34
-------
BOD test has been used to approximate carbonaceous matter In waste
waters (Van Hall and Stenger, 1964). However, new TOG analyzers that
provide a rapid and simple means for determining organic carbon levels
in aqueous samples are now available (Van Hall and Stenger, 1967).
A Beckman Model 915 TOG Analyzer was used in this study. There
were no operational problems except for the optional air purification
unit which proved unsatisfactory because of a failure of the diaphragm in
the air pump after a short period of use. This caused improper flow
and pressure conditions to exist in the analyzer and necessitated the
use of a bottled source of zero grade air.
The syringe used for measuring and injecting the sample into the
TOG analyzer has a needle opening of 170y in diameter and the sample
volume is small, generally 20ul. Thus, blending is necessary for samples
containing suspended material in order to prevent the screening out of
large particles and to assure uniformity of samples (Schaffer et al,
1965). Since the peak value depends on the rate of sample injection as
well as the carbon content of the sample, the rate of injection must be
uniform for each determination.
Sample dilution with carbon-free water is necessary when carbon
concentrations are greater than 4000 mg/1 (maximum range of analyzer)
or when high contents (10,000 mg/1) of salt, acid, or base anions are
present (Ford, 1968a). A high anion content interferes with the infrared
energy absorption pattern.
Nutrient Analyses
Because of delays in receiving the Auto-Analyzer equipment and of
heavy laboratory workloads, most samples had to be stored a month or
more before nutrient analyses were performed. Some were stored as
long as six months before all analyses were made. At first, mercuric
chloride, HgClo, was added to the samples (Jenkins, 1967) followed by
refrigeration for preservation. The method was abandoned because the
HgCl? caused a precipitate to form in the lagoon samples. This tended
to clog the tubing of the Auto-Analyzer. Freezing was used on all other
samples from the swine sites and proved to be satisfactory although not
ideal. Five ml of chloroform was added per liter to all samples from
the dairy, beef, and poultry sites before freezing the samples. Three
portions of each sample were frozen so the tests could be performed on
newly thawed samples. Some samples were refrozen and later thawed
and tested again. There were occasional apparent changes in nutrient
contents of samples without chloroform handled in this manner. Also,
when the lagoon samples were thawed, insoluble precipitates not observed
before freezing were present.
Commonly used Auto-Analyzer methods for determining TN, N02, N03
and OP proved quite reliable. TP and NH3 analyses of the lagoon wastes
presented difficulties until appropriate methods were found to handle
samples with considerable particulate matter. A method using a continuous
digestor gave good TP results. A dializer was used in the NH3 determin-
ation to eliminate the solids problem. This technique gave good results.
35
-------
In all tests alternate samples of water were used for washout of the
flow system. This gave an effective rate of sampling of one-half the
recommended rate. Every tenth sample was a standard. All Auto-Analyzer
determinations conformed to procedures described in Standard Methods^
(APHA, 1965). Checks made against the manual determinations gave the
same results within the expected accuracy of each test.
Total Nitrogen. TN, as used herein, is total Kjeldahl nitrogen.
It is the sum of the free ammonia and organic nitrogen compounds that
are converted to ammonium sulfate under the conditions of digestion
which are specified below.
Basically, the quantitative determination of TN was per Technicon
Corporation (1965) Bulletin N-3 and the FWPCA (1969) automated phenolate
method. It involved complete digestion of organic material followed by
measurement of the quantity of ammonia produced. The samples were auto-
matically digested using an Auto-Analyzer continuous digester and a
sulfuric acid solution containing perchloric acid and vanadium pentioxide
as a catalyst to convert organic nitrogen to (NH^^SO^. The solution
was then automatically neutralized with sodium hydroxide and treated with
alkaline phenol and sodium hypochlorite. This treatment formed a blue-
green color designated as indophenol and was measured colorimetrically.
Sodium nitroprusside was used to increase the intensity of the color
for samples containing low levels of TN (FWPCA, 1969. Brij-35 detergent
was added to the alkalin phenol reagent at the rate of one-half ml/1 to
improve the flow patterns.
Ammonia. The Auto-Analyzer procedure used to determine NH-j concen-
trations in the samples was the alkaline phenol-hypochlorite method
proposed by O'Brien and Fiore (1962). Brij-35 detergent was added to
the sodium carbonate and sodium phenate reagents at the rate of one-half
ml/1 to improve the flow patterns and prevent plugging of the flow
system. Methods not employing a dialyzer were not as satisfactory as
the one used. The samples usually contained a large amount of particulate
matter even after filtering with a continuous filter. This caused
erratic flow patterns in the system and poor results.
Nitrite-Nitrate. The Auto-Analyzer procedure used to determine
NC>2 and N03 contents of the samples was similar to the adaptation of the
diazotization method proposed by Kamphake et al, (1967). Under alkaline
conditions the N02 ions reacted with sulfanilamide to yield a diazo com-
pound that coupled with N-1-naphthylethylene-diamine dihydrochloride to
form a soluble dye which was measured colorimetrically. All N(>3 in a
sample was first reduced to N02 by an alkaline solution of hydrazine
sulfate containing a copper catalyst for N02 plus NO^ measurements.
Subtraction of the N02~N originally present in the sample from the total
N02-N found after reduction gave the original NOg-N concentration. One-
half ml of Brij-35 detergent was added to each liter of color reagent to
improve flow patterns. When NOo was not measured, the copper reagent
and hydrazine sulfate inlet lines were placed in water and the heating
bath was by-passed.
Total Phosphate. The procedure used to measure TP contents of the
samples first converted the phosphates present to orthophosphate by using
36
-------
the Kjeldahl digestion procedure reported above. The phosphate concen-
tration was then determined by the reduction of phosphomolybdic acid
with aminonaphtholsulfonic acid (ANSA) as proposed by Lundgren (1960)
for determining total inorganic phosphate. The Lundgren method for con-
verting the condensed phosphates present to orthophosphate by means of
hydrolysis with sulfuric acid was not generally satisfactory because the
samples often contained a large amount of particulate matter which caused
erratic flow patterns and because a total phosphate rather than a total
inorganic phosphate analysis was desired. About 300 samples were analyzed
using both digestion procedures. The values were approximately the same
for both digestion procedures except when the Lundgren method showed
erratic results. The latter was probably due to poor flow patterns
caused by particulate matter. The TP data reported herein were obtained
using the Kjeldahl digestion method. Levoir-4 detergent was added to
the ANSA at the rate of one-half ml/1.
Orthophosphate. The Auto-Analyzer procedure used for the determina-
tion of OP employed the reaction of ammonium molybdate in an acid
medium to form molybdophosphoric acid that was then reduced to the
molybdenum blue complex by reaction with aminonaphtholsulfonic acid
reagent. Technicon Corporation (1969) Industrial Method 2.68W was used.
One-half ml of Levoir-4 detergent was added to each liter of ANSA to
improve the flow pattern.
37
-------
EXPERIMENTAL DATA AND RESULTS
Introduction
Interpretations of the results of the study to ascertain the amounts
of and factors governing pollutional loadings reaching streams from
animal production areas are presented in this section. Only sample data
and results are presented herein. Full presentation of data, data tables,
and figures are given in the Supplemental Appendix available at the
Environmental Protection Agency, the Department of Biological and Agri-
cultural Engineering at North Carolina State University, the Water
Resources Research Institute of the University of North Carolina (Raleigh),
and the North Carolina Department of Water and Air Resources.
Experimental Data
Corresponding hydrological and effluent quality data from all sites
are printed by data groups and graphed against time in the Supplemental
Appendix. With the exception of NOo, from sites A, B, and C and COD for
the seven swine sites, the following variables are plotted for all sites:
rainfall, qt (reported as depth of flow), Tx, Tn, FC, TC, COD, BOD, TOG,
TN, NH3, N03, TP, and OP. No TC or OP data is reported for the dairy,
beef, and poultry sites. Conductivity, pH, FS, TS, VS, and COD data for
the swine sites are also presented by site and time of collection.
When more than one value of a pollution index was available for a
given time because of collection of multiple samples, and/or determina-
tion of a particular index more than once on a given sample, average
values of the indices are given. Missing index data resulted from either
analytical failure or, more commonly, from a decision not to perform the
analysis because of insufficient need.
N02 and NO-^. The N03 values reported herein are the sum of N02 and
N03. ' N0£ levels in all samples and N03 and N02 in the lagoon effluent
samples were not significant and, thus, are not reported. The combina-
tions of N02~N plus NO«-N averaged only 0.3 mg/1 and never exceeded 0.5
mg/1 for more than two hundred samples from the three lagoons. The
N03-N/N02-N ratio for all lagoon samples was about 0.10 and for the
stream samples was approximately 10.
The BODc data deserve elaboration because the results of the analyses
for the more polluted samples (BOD5 greater than 60 mg/1) were generally
dependent upon the dilution factor, as shown by the following representa-
tive results.
0.75 590
1.00 510
1.50 470
39
-------
The BOD variations for strong wastes were discovered early in the study,
and a thorough review of procedures failed to account for the variations
in terms of procedural errors. Thus, the variations were attributed to
characteristics of the BOD test and/or the wastes. Clark (1965)
reported that the BOD test was not applicable to swine waste lagoon
waters and attributed this to antibiotics, etc., that retarded bio-
degradation of the wastes. A study of feedlot wastes by Morrison et al
(1969) suggested that antibiotics could be present in animal wastes at
sufficient levels to cause problems with bio-assays. Studies by
Taiganides and White (1969) pointed out limitations and problems asso-
ciated with applying the BOD test to swine wastes. Scaffer ejt al (1965)
reported the phenomenon of "sliding BOD" for municipal wastes, i.e., lower
BOD results with larger amounts of sample. Long-term BOD results for
strong wastes were subject to similar variations. The long-term BOD plot
in Figure 11 is representative of data obtained from twelve effluent
samples analyzed from the three lagoons. The twenty-day BOD values were
generally twice the 6005 values, and the rate of increase with time was
still substantial after twenty days with no indication of onset of a
nitrification phase in the classical sense.
A total of 35 samples, 23 of which were lagoon effluent samples,
were seeded for BOD analyses. About one-half of the samples were
inoculated with a seed developed from lagoon effluent while the rest
were inoculated with seed developed from domestic sewage. Standard
seed development and inoculation procedures were followed. There were
no observed differences between seeded and unseeded samples for either
BOD,, or long-term BOD.
The BOD data reported herein are from analyses in which more than
2 mg/1 of oxygen was utilized and more than 1 mg/1 of oxygen remained
after incubation. When variations occurred, the BOD value obtained from
the less dilute samples was taken as the correct value except when two
or more values for other dilutions were in close agreement but different
from this value. In the latter instance, the BOD in best agreement with
the less dilute sample was chosen. This procedure produced results
consistent with findings for other indices. An intensive review of
alternative methods showed no improvement over this method of selection.
Simple Statistics and Regression Analyses
The data collected during the study were subjected to standard com-
puterized statistical analysis in order to determine the applicability of
the procedures used in the study, to evaluate the suitability of the
pollution indices, and to develop relationships for estimating the
effects of animal production units on water quality under similar manage-
ment conditions. Simple regression analyses revealed relatively high
correlations between many of the pollution indices. These correlations
suggest that substantial savings can be made in stream quality surveys
by limiting the analyses to a few key indices with the remainder being
estimated.
Means (M), standard deviations (s), and linear correlation coeffi-
cients (r) for the data are presented in Tables 27-80 of the Supplemental
Appendix. The number of values, No., used in each analysis is also
40
-------
1000
800
600
t»o
o
m
400
200
BODon = 760 mg/1
BOD = 50 mg/1
24
Figure 11. Representative long-term lagoon effluent BOD
41
-------
recorded. The simple statistics reported in Tables 41, 54, 67 and 80 were
obtained using all values of data for the given data group; i.e., FC
for data group A in Table 41 is the mean for all FC values from site A.
The simple statistics reported in the remaining tables were obtained
using paired observations. The means for some indices are larger than
the means of related indices_in apparent contradiction to what might
normally be expected; e.g., FC larger than TC. This is due to the
fact that the TC and FC data are not necessarily mutually paired when
paired with other indices.
Although the data were grouped to make analyses and presentation as
meaningful as possible, any interpretations of the results reported must
be made with due caution and a full understanding on the particular data
bit. For example, q for data group E in Table 41 is simply the average
of the individual q values recorded at the 176 times of sampling at
site E. Since sampling was commonly associated with rainfall-runoff
events, qt does not represent the average flow rate during the study
period.
The data obtained during this study are random pairs of observations
from bivariate normal distributions. Thus, the regression equation of
y on x is:
s
(x - 30
Similarly,
s
X
s
x
= xH-r_- (y _ y)
y
When the correlation between two indices is high, these equations may be
used under comparable conditions to obtain equations for estimating the
value of one indice given another.
For example, r for FC vs TC for data group E, i.e., all paired FC
and TC observations from site E, is 0.957. High r values are also
reported for sub-groups of data group E. Thus, the need for both
determinations in similar studies is nonexistent since one index can be
determined from the other with about 92 percent accuracy. If FC were
selected as the indice to be determined by analyses, the prediction
equation for TC from land disposal sites similar to site E would be:
TC = 39,540 + 0.957 ' (FC - 19,050)
= 2.1 FC + 100
The TOG test proved particularly promising for use in developing
relationships to evaluate other indices because it is simple, rapid, and
42
-------
reliable and correlates well with the other indices. Certain correla-
tions between TOG and the other indices are summarized in Table 1.
These are all significant at the 0.01 level except for FC from sites X,
J and ABC.
Table 1. Correlation coefficients, r, for TOC with other
pollution indices.
Site
Control, F
Land Disposal
E( swine)
K(swine)
P (poultry)
X (poultry)
Z(beef)
Lagoons , ABC
Direct Discharge
D( swine)
H (dairy)
J (dairy)
r
FC
.406
.437
.794
.491
.050
.895
.037
.655
.838
.207
BOD5
.734
.966
.920
.610
.972
.724
.666
.955
.970
.893
TN
.943
.875
.665
.926
.955
.637
.676
.708
.583
TP
.883
.899
.425
.730
.620
.473
.735
.714
.393
The following simple regression equations with TOC as the independ-
ent variable and BODc as the dependent variable demonstrate the types of
potentially useful relationships which can be developed for the assess-
ment of pollution under applicable conditions:
Land runoff from:
Swine wastes
BOD5 = 0.79 TOC - 2.5
Dairy wastes
BOD5 = 0.66 TOC - 2.0
Poultry wastes
Site
E
H
P & X
BODC
0.35 TOC
Multiple RegressionAnalyses for Data from the Swine Sites
Results of multiple regression analyses of the swine waste data with
FC, TC, BOD5, TOC, TN, NH3> N03> TP, and OP each as the dependent variable
43
-------
and Tn, H, and q0 as the independent variables are presented in Appendix F
of the Supplemental Appendix. The coefficients of determination, R2, an
-------
2
Nevertheless, the high R values for some of the equations obtained
for the land runoff data groups (sites E, F, and K) show promise that
equations of general applicability for predicting stream pollution from
land runoff sites might be developed with a more detailed and longer
term study. Estimating equations for BOD5 and TOC from the control
watershed (F) are examples. Normative values for these indices ranged
between 0-2 and 5-10 mg/1, respectively, in the absence of rainfall.
The regression equation from data collected between March 18-25, 1969,
BOD5 - 2.8 - (0.035) T + 2.2 (q )
2
has an R value of 0.604 and produces a BOD5 estimate of 1.1 mg/1 at a
minimum daily temperature of 50°F and a runoff flow rate of zero. The
TOC regression equation for the same data
TOC = 5.6 + (0.024) T + (13.9) q
n o
2
has an R value of 0.498 and estimates a TOC of 6.8 mg/1 under the same
conditions. At a runoff flow rate of 1 cfs and Tn of 50°F, the equations
estimate a BOD5 of 3.3 mg/1 and a TOC of 20.7 mg/1. These values are
fairly consistent with normative values observed during runoff periods.
In other multiple regression analyses, q was replaced by other flow
data; e.g., qt, q^, and qt to various powers, and Tn was replaced by TX.
These attempts did not provide an improvement in the ability to predict
the pollution indices and, thus, the results are not reported.
Selected multiple (and step-wise) regression analyses were performed
using BOD5, TOC, etc. as independent variables. Examples of these
analyses are reported in Appendix G of the Supplemental Appendix.
Although these analyses generally gave considerable improvement in the
resulting regression equations for the land runoff sites, their use
requires that a pollution indice, e.g., TOC, for the wastewater be known.
For example, the BOD^ regression equation for data group E with Tn, H,
and qo as the independent variables
BOD. = -1.3 - 0.517 T + 0.012 H + 34.1 q
5 n o
2
has an R value of only 0.435 while the BOD5 equation with TOC added as
an independent variable
BOD,, = 2.1 + 0.9 TOC - 0.05 T - 0.004 H - 6.4 q
5 n o
has an R2 value of 0.946. The equation with only TOC and qQ as the
independent variables
BOD5 = 3.3 + 1.08 TOC + 10.1 qQ
has an R2 value of 0.947. These equations indicate that T and H were
not important in determining the BOD5 in the stream under the conditions
and within the time span experienced in this study. They also substantiate
45
-------
the high correlations between TOG and other pollution parameters
developed in the simple regression analyses.
The step-wise equations for the lagoon effluent data still showed
low R2 values and inconsistency. This reflects the low correlations
between variables in these data. Although a number of the equations for
the direct discharge stream show high R2 values, they are still unsatis-
factory as prediction equations in light of the changes which occurred
in pollution indices with time and with the intensity of pollution which
are not accounted for by the equations.
Swine Waste Studies
Lagoon Influent Studies
Values of the pollution indices obtained from samples of the
influent (diluted fresh swine waste) to the lagoons were quite variable.
Based on only 10 samples, typical indice values for the influent to
lagoon A were:
TC 20 x 106 colonies/100 ml
FC 6 x 106 colonies/100 ml
FS 1.5 x 106 colonies/100 ml
BOD5 725 mg/1
TOG 680 mg/1
COD 1400 mg/1
TS 1700 mg/1
VS 1000 mg/1
TN-N 200 mg/1
NH3-N 100 mg/1
N03-N 1 mg/1
TP-P04 85 mg/1
OP-P04 75 mg/1
Higher values were common for influent samples to the other lagoons due
to lesser amounts of dilution. Not enough influent data were obtained
to accurately delineate the amount of waste contributed to the lagoons
per hog per day. However, when the number of hogs and volume of wash
water were considered, the 10 samples collected at site A indicated that
the daily production per hog of FC, BOD^, COD, TS, TN, and TP was within
the range reported by others for fresh swine waste; e.g., Muehling (1969).
The influent FS levels were much lower than expected, and the FC/FS
ratio was normally four for all lagoon influents. Thirty subsequent
samples of the lagoon influents verified this ratio. According to
Geildrich (1966) the expected ratio for waters containing animal wastes
is less than 0.5. The higher FC/FS ratio reported herein may have been
caused by the KF medium used in the FS test. The medium was supposed
to inhibit growth of bacteria other than FS and to produce colonies
light to dark pink in color. However, pale yellow colonies were normally
present after incubation. Only the pink colonies were counted as FS.
46
-------
Lagoon Effluent Studies
The three lagoons investigated in this study had surface outflows
that were measured and analyzed to determine the amount of water pollu-
tion resulting from anaerobic lagooning operations. Any water pollution
via subsurface discharge into the streams from the lagoons was not
measured. This was judged minimal because the lagoons were located
in clay soils that had low hydraulic conductivities and because of the
sealing properties of lagoon liquor. In fact, subsurface flow into
lagoon A was more of a possibility as discussed later.
Examples of effluent quality data and corresponding hydrological data
for lagoon A are presented graphically in Figures 12-18» The large
variations encountered in most of the data from the lagoons emphasize the
need for caution when only occasional samples are used to characterize
the water pollution potential from this source.
The number of unit hogs,, H, at each site were determined by divid-
ing the total weight of swine in pounds by 100 Ibs. H remained nearly
constant at each lagoon during the study periods. The slight variations
(± 101) that did occur were judged insignificant for the study. The
number of unit hogs for sites A, B, and C were 300, 100, and 90, respectively.
Factors such as the amount of wash water used per animal markedly
affected the concentrations of pollutants between sites. Thus, the
effect of H on the pollutions! yields from the lagoons presented herein
was evaluated in terms of the amounts of pollutants discharged in the
effluents per H as opposed to relating the variations in concentrations
of pollutants to H.
Hydrological Data. Rainfall in the area of the lagoons averages
50 in./yr. and evaporation for small lakes in the geographical location
is normally 35 in./yr. Evaporation from lagoons was not investigated,
but was probably less than that from small lakes because of surface
solids and scums. Even if evaporation from lagoons were equal to that from
small ponds, a lagoon would produce an effluent because of the 15 surface
inches per year of rainfall in excess of evaporation, the waste volume,
the surface runoff from surrounding areas, and the wash water. Subsurface
inflow and/or outflow would also be considered in a water balance. The
contribution due to waste volume would generally be insignificant in a
water balance. The production of fresh swine wastes is about 1 gal/day/H.
As exemplified in Figure 12, rainfall produced hydrographs of sur-
face effluent flow from the lagoons similar to hydrographs characteristic
of a small watershed, i.e., a rapid increase in flow rate followed by a
prolonged die off to base flow. Wash-down events were not discernible
from effluent flow records at site A. Small peaks in the flow rate were
identified at site C when wastes were added to the lagoon. Occasional
increases in the rate of effluent discharge were observed at site B
following wash-down operations. Some of the variations in effluent flow
rates from the lagpons were caused by floating debris that restricted
the flow through the overflow pipes. The variations in qt at site A
during November 20-23, 1968,, were accompanied by an apparent inversion of
the lagpoa liquor caused by a temperature change.
47
-------
00
TIME OF
ITEAK FLOW
80 0
TIME, HOURS
80
Figure 12. Hydrological Data, Site A, Oct. 19-21 and Nov. 20-23, 19.68.
-------
3.Qr-
2.5 -
K
!\
I \
I
l\
I \
I \
I \
I \
TOTAL COLIFOBM
——— FECAL COLIFORM
o
o
a
M
*•
N
2,0
1.0
0.5
1 !
II M
M I
"I
I I
I
I
\/
TIME OF
IPEAK FLOW
i
\
\
20
40
\ I1 '
\ S '
XI I
it
i
I
I
(TIME OF
• *PEAK FLOW
60
80 ' 0
TIME, HOURS
20
40
60
80
Figure 13. Bacteria Counts, Site A, Oct. 19-21 and Nov. 20-23, 1968.
-------
600
BOD
TOG
500
400
m
o
300
200
100
I
TIME OF
PEAK FLOW
20
40
60
1
80 ' 0
TIME, HOURS
20
40
1
TIME OF
PEAK FLOW
60
80
Figure 14. BOD5 and TOG, Site A, Oct. 19-21 and Nov. 20-23, 1968.
-------
AOOr-
KJELDAHL
300
200
H
H
AMMONIA
100
TIME OF
PEAK FLOW
TIME OF
PEAK FLOW
I K I
I
20
40
60
80 ' 0
TIME, HOURS
20
40
60
80
Figure 15. Kjeldahl and Ammonia Nitrogen, Site A, Oct. 19-21 and Nov. 20-23, 1968.
-------
200
TOTAL
ORTHO
150
O
Pk
CO
a
100
to
o
PS
50
TIME OF
PEAK FLOW
I
I
I
I
I
TIME OF
EAK FLOW
20
40 60 80 ' 0 20 40
TIME, HOURS
Figure 16. Phosphate, Site A, Oct. 19-21 and Nov. 20-23, 1968.
60
80
-------
Ui
LO
0.6
S5
CO
0.4
oo
e
0.2
NITRITE PLUS NITRATE
NITRITE
TIME OF
PEAK FLOW
I
I
I
20
40
\
I
I
I
TIME OF
PEAK FLOW
I
I
60
20.
4Q
60
80 Q
TIME, HOURS
Figure 17. Nitrite and Nitrate Nitrogen, Site A, Oct. 19-21 and Nov. 20-23, 1968.
80
-------
2500
TOTAL
VOLATILE
2000
Ul
1500
S
JM
t-1
o
en
1000
V V -
-\ r
v
500
I
TIME OF
PEAK FLOW
TIME OF
PEAK FLOW
I
20
40
60
80 0
TIME, HOURS
20
40
60
80
Figure 18. Solids, Site A, Oct. 19-21 and Nov. 20-23, 1968.
-------
Neither qt, qb, nor qo correlated well with the pollution indices
for most lagoon effluent data groups. The poor correlations might be
explained by noting that the lagoon volumes provided ample opportunity
for the lagoon liquor to reach uniformity with respect to pollutional
content and any variation in flow simply removed more or less liquor
that had a rather constant waste content. Thus, other factors were more
important in governing the pollution indice values than was flow rate.
The largest surface effluent discharge rates occurred during late
fall to early summer when rainfall was greatest and evaporation least.
The effluent discharge results of Table 2 were obtained by assuming
that the average effluent flow rates which occurred at time of sampling;
i.e., q. represents the average flow rate (q).
Table 2. Rates of effluent flows from the lagoons
(Site)
A
B
C
a*, cfs
max
.285
.057
.052
mean
.027
.002
.002
min
.001
0
0
__
*l
ac-f±
yr
19.2
1.6
1.2
ft3/yr
840,000
68,000
50,000
gal /day
17,250
1,400
1,050
gal/day/H
58
14
12
The large yields from lagoon A suggest subsurface flow into the
lagoon. The measured rate of wash-water usage in the operation was
15 gal/day/H. An estimated additional 15 gal/day/H was added to the
lagoon by leaky water valves and swine wastes. A difference in rain-
fall and evaporation of 15 in./yr. represents a net yield of 4.5 gal/day/H
from the 50,000 ft2 lagoon. An assumed yield of 30 in./yr. of surface
runoff entering the lagoon from the surrounding 20,000 ft2 contributed
another 3.5 gal/day/H. A large portion of the 20 gal/day/H not accounted
for must have come from subsurface sources. This is plausable because
the lagoon was the headwater of a continuously flowing stream.
The effluent yields from lagoons B and C were in the range expected.
Wash-water usage was 10 gal/day/H and waste volume was estimated to be
1 gal/day/H. The 15 in./yr. difference in rainfall and evaporation
accounted for an additional 1.8 gal/day/H. Surface runoff at 30 in./yr.
accounted for 1.7 gal/day/H at site B and 0.8 gal/day/H at site C for
totals of 14.5 and 13.6 gal/day/H, respective.y
Bacteria Indices. Bacteria counts in the effluents from the three
lagoons were generally highest for lagoon B and lowest for lagoon C.
The trend shown in Figure 13 during the October 19-21 period for TC and
FC counts to increase and decrease in direct relation to the effluent
hydrograph was not verified by other data. In fact, the correlation
coefficients indicate almost no general linear relationships existed
between the bacteria indices and the other pollution indices or hydrologi-
cal factors. However, the bacteria indices were well correlated with
55
-------
each other at sites A and C.
at site B.
Usually, TC and FC were poorly correlated
Bacteria counts were lower during the late spring and summer period
than during the fall and winter months. This is reflected by the nega-
tive correlation coefficients for FC and TC with Tx and TR for data
obtained from sites B and C. However, the failure of such a general
correlation between bacteria counts and temperature for data from site A
suggests that temperature was not the sole factor involved in this
phenomenon.
The lower than normal TC values and TC/FC ratios obtained during
November 21-23 at site A (Figure 13) and on November 23, 1968, at site B
may have been caused by higher than normal concentrations of antibiotics
and/or other toxic substances in the lagoon liquor if fecal organisms are
assumed to be more tolerant to these substances than are total coliform
organisms. Erratic variations in BOD5 levels also occurred at lagoon A
during this period. This suggests that an inversion of lagoon liquor
caused by the drastic drop in temperature occurred. This supposed
inversion of lagoon liquor could have released toxic materials from the
lagoon bottom. Clark (19) reported a similar bacteria ratio disturbance
in swine lagoon waters receiving wastes from hogs that had been fed
unusually large amounts of antibiotics plus sulfur drugs.
Results of calculations of the bacteria yields in the lagoon
effluents are shown in Table 3. The value for FC/day/H for data group A
was obtained by multiplying the mean of all FC measurements made on the
effluent samples from lagoon A by q" for data group A from Table 2. The
value of TC/FC for data group A was obtained from all paired TC and FC
measurements made on the samples from lagoon A. The other results were
obtained in a similar manner. The average yields or FC were only 10
percent at site A, 6 percent at site B, and 1 percent at site C of the
average production rates for fresh swine wastes, 8.9 x 10^ col/day/H,
reported by Gieldreich (1966).
Table 3. Bacteria in effluents from the lagoons
GROUP
(Site)
A
B
C
FC
106 col/ 100 ml
max
L.05
2.10
.97
mean
.42
1.04
.17
min
.03
.06
.01
106
col/day/H
900
490
90
TC
106 col/ 100 ml
max
3.00
4.10
1.45
mean
1.22
2.06
.38
min
.22
.30
.01
106
col/day/H
2,600
1,000
200
TC/FC
2.9
2.0
2.2
BOD5 and TOC. The concentrations of BOD5 in the lagoon effluents and
the total amount of BOD5 discharged to streams were usually greater during
the colder months than during the warmer periods. This may have been
partially due to the more favorable temperature for biological activity
56
-------
during the warmer periods and thus more biological treatment of wastes.
However, the correlation coefficients between BODr and temperature were
generally too low and inconsistent to attribute this phenomenon solely
to this hypothesis.
TOG values for the lagoon effluent were quite reproducible and free
of noticeable errors. The test was free of the erratic variations
common to the BOD test. The data collected during the study showed that
the TOC values followed the same general pattern of increasing during
the colder periods and decreasing during the warmer months as did the
BOD5. Variations in TOC with season were generally less than those for
During the early part of the study, BOD5 values generally exceeded
TOC values. After February 1969 TOC values generally exceeded BODs
values for site B. This change was delayed until April 1969 at sites A
and C. A reversal (TOC greater than 8005) did occur at site A during
the period of November 21-23, 1968 (Figure 14). As discussed previously,
an inversion of the lagoon liquor probably occurred during this time due
to a drop in temperature. If toxic substances were released from the lagoon
bottom, these could have retarded the biochemical oxygen demand, reducing
the BOD5/TOC ratio.
The reversal in the BOD5/TOC ratio during the spring for all three
lagoons might have been caused by an increase in toxic materials in the
lagoon waters during the warmer periods, but no justification of this
hypothesis is available. Another possible explanation of the reversal
is that the wastes were more highly oxidized during the warmer period.
Since the decrease in BOD5 from winter to spring was more drastic than
that for the TOC, this explanation seems more plausible.
TOC and 6005 values in the effluents from the three lagoons were
generally highest for lagoon C and lowest for lagoon B. This was just
the opposite of the bacteria values which were highest for lagoon B and
lowest for lagoon C.
Neither 8005 nor TOC values for the lagoon effluent correlated well
with the other pollution indices. TOC was only slightly better than BOD5
in this respect.
Results of calculations of the BOD5 and TOC yields in the lagoon
effluents are shown in Table 4. The value for BOD5/day/H for data
group A was obtained by multiplying the mean of_ all BOD5 measurements
made on the effluent samples from lagoon A by q for data group A from
Table 2. The value of BOD5/TOC for data group A was obtained from all
paired BOD5 and TOC measurements made on the samples from lagoon A. The
other results were obtained in a similar manner. Although large differ-
ences in BOD5 and TOC occurred between individual samples, the results
are similar in terms of Ib/day/H. Thus, in the long run the same con-
clusions concerning the applicability of lagoons for swine waste treat-
ment can be drawn from either the BOD,, or the TOC.
57
-------
Table 4. BOD,, and TOC in effluents from the lagoons
/""Tl/^TTD
GROUP
(Site)
A
B
C
ROT)
mg/1
max
610
600
1015
mean
390
270
610
min
90
80
215
Ib/day/H
.181
.028
.071
TOC
mg/1
max
510
550
815
mean
360
340
470
min
165
200
355
Ib/day/H
.170
.035
.055
BOD5/TOC
1.1
0.8
1.4
The BOD5 in Ibs/day/H in the effluents were 59 percent at site A,
9 percent at site B, and 22 percent at site C of the average production
rates for fresh swine wastes, 0.32 Ib/day/H, reported by Meuhling (1969).
The apparent differences in treatment efficiencies between lagoons may
be partially attributed to the differences in feeding practices. As
mentioned earlier, a particularly high rate of copper sulphate was fed
at site A and cooked garbage was a major part of the feed. The opera-
tions at sites B and C followed recommended feeding practices.
A more plausible explanation is that 6005 and TOC yields were
related to the time of waste retention. The volumes of lagoons A, B,
and C were about 250, 120, and 160 103ft3 while the effluent flow rates
were 840, 68, and 50 103ft3/yr, respectively. The volume displacement
times obtained by dividing the effluent flow rates into the volumes were
0.3, 1.8, and 3.2 yrs, respectively. These figures must be interpreted
in a relative rather than absolute sense, however, in view of the flow-
through measurements which demonstrated that a portion of liquid swine
wastes flowed through the lagoons in a few hours with the major part
being retained for longer periods.
The differences in age of the lagoons may have been an added factor
in governing the 8005 and TOC yields. The lagoons had been used 6, 3,
and 2 years, respectively, before the study began. Solids had built up
above the surface at the inlet of lagoon A. No solids were observed
from the banks of the other two lagoons.
Differences in capacity/H may have been another factor. The values
were about 830, 1200, and 1800 ft3/H while surface areas were about 170,
200, and 220 ft2/H, respectively.
Any one of the above factors may have influenced the BODc and TOC
yields. The most plausible factor is retention time because it showed
sufficient variation to account for the large variation in the indices,
and it is related to the other factors involved.
Nitrogen. Total (Kjeldahl) nitrogen concentrations doubled at
sites B and C from the fall of 1968 to the spring of 1969. They were
highest in early May with values of 280 and 560 mg/1, respectively.
During the same period, TN concentrations decreased at site A and con-
tinued to decrease until the end of the study, summer of 1969. The poor
58
-------
correlations between TN and qt suggest the changes in concentrations
were not caused by variations in flow and/or dilution by rain water.
TN and NH3 were linearly related to each other. However, the poor
coorelations for some data groups preclude recommending substitution of
one test for the other. Ammonia nitrogen levels followed the trends of
TN and accounted for 60-100 percent of the TN. The NH3 averaged 80 per-
cent of the TN for all lagoon effluent samples. Correlations between
TN and NH3 and the other data were not consistent and were often
nonsignificant.
Results of calculations of the nitrogen yields in the lagoon
effluents are shown in Table 5. The value for N/day H for data group A
was obtained by multiplying the mean of all TN measurements made on the
effluent samples from lagoon A by q" for data group A from Table 2. The
value of TN/NH3-N for data group A was obtained from all paired TN and
NH3 measurements made on the samples from lagoon A. The other results
were obtained in a similar manner. At a land application rate of 1/10
ft of lagoon liquor the nitrogen application rate would be 66, 46, and
78 Ib/acre from lagoons A, B, and C, respectively.
Table 5. Nitrogen in effluents from the lagoons
GROUP
(Site)
A
B
C
N
mg/1
max
330
345
560
mean
244
168
287
min
120
88
205
Ib/ac-ft
660
450
780
Ib/day/H
.115
.018
.033
TN/NH.-N
3
1.2
1.4
1.2
The results from sites B and C suggest that some nitrogen may be
removed by lagoons. This was probably associated with solids removal
and/or retention.
Phosphate. Total phosphate concentrations in the effluents were
rather uniform at each site throughout the study periods, but variations
between samples collected at close time intervals were greater than
expected. There was a slight decrease in TP content at each site during
early spring. This was followed by a slightly larger increase at sites B
and C while at site A the TP level returned to its previous value.
Generally, TP and OP were linearly correlated. OP averaged 83 per-
cent of the TP for all lagoon effluent samples. Correlations between TP
and OP and the other data were often non-significant.
59
-------
Results of calculations of phosphate yields In the lagoon effluents
are shown in Table 6. The value for PO^/dayjR for data group A was
obtained by multiplying the mean of all TP measurements made on the
effluent samples from lagoon A by q~ for data group A from Table 2. The
value for TP/OP for data group A was obtained from all paired TP and OP
measurements made on the samples from lagoon A. The other results were
obtained in a similar manner. At a land application rate of 1/10 ft of
lagoon liquor the P205 application rate would be 29, 22, and 24 Ib/acre
from lagoons A, B, and C, respectively.
Table 6. Phosphate in effluents from the lagoons
O'D/^TT'D
LiKUUjr
(Site)
A
B
C
P04
mg/1
max
200
214
160
mean
142
112
116
min
74
78
80
Ib/ac-ft
390
300
320
Ib/day/H
.069
.012
.013
TP/OP
1.2
1.2
1.2
Discussion. Although large variations in pollution indice values
were encountered between samples and lagoons and with time, sufficient
data were collected to assure that the results are representative of
normal contributions of anaerobic swine waste lagoons in the Piedmont
area of North Carolina. While the study lagoons provided more capacity
per animal than USDA and other recommended standards, the lagoon
effluents clearly exceeded raw domestic sewage in strength.
The study lagoons appeared to function mainly as traps, i.e.,
settling and retention basins, and provide only a limited amount of treat-
ment beyond that experienced through sedimentation. Lagoons B and C
functioned more efficiently with regard to amounts of pollutants removed
but not in terms of the strength of the effluents. At site A the calculated
decreases in bacteria and organic pollution discharged compared to the
estimated inputs based on characterization of fresh wastes must be
viewed in light of the corresponding increases in phosphorus and nitrogen.
Such management practices as wash-down were of major importance in
controlling the amount of wastes reaching streams. The waste load to
receiving streams increased with volume of effluent and decreased with
longer periods of waste retention.
Direct Discharge Study
The direct discharge site was studied only a short period of time
because the practice of directly discharging wastes into streams was
judged unacceptable from the beginning and the need for data was limited
60
-------
to demonstrating the impact of this type of animal waste disposal on
surface water quality. Examples of effluent quality data and correspond-
ing hydrological data from site D are presented graphically in Figures
19-24.
The stream bed at site D was relatively free of deposited solids
during the early part of the study. Shortly thereafter, however, it
was normally covered as deep as three inches with solids from swine
wastes. These solids were primarily from wastes washed directly into the
stream from the farrowing house and holding pens.
The number of animal units, H, was obtained by dividing an esti-
mated total weight of swine in the farrowing house, the pens, and the
drylots by 100 Ibs. Since the swine wastes in the drylots remained
away from the stream except during rainfall-runoff events while the
wastes from the farrowing house and holding pens were washed directly
into the stream every two days, the H values reported are somewhat
lacking for a basis, Nevertheless, the significant linear correlation
coefficients between H and the pollution indices suggests that the H
values were important in determining the pollution indice values and
loads on the stream.
Hydrological Data. Temperature variations during the study were
small. This may account for the low correlation between the temperature
and the pollution index values.
The stream flow hydrographs for site D (Figure 19) show peaks caused
by washing wastes from the farrowing house and holding pens in addition
to the normal responses to rainfall. The periods of wash-down are marked
on the plots to the extent possible. Some periods of wash-down may have
been overlooked.
Base flow rates (q^) correlated better with the pollution indices
than either qt or qo. The qt value for data group D, obtained by averag-
ing the flow rates at the times of sampling, was higher than the actual
mean flow rate because the sampling times were commonly associated with
surface runoff and wash-down events. The flow results presented in
Table 7 were obtained by using ~q^ as the mean flow rate.
Pollution Indices. Indicator bacteria counts were exceptionally
high in the stream samples from site D. The minimum FC value observed
was obtained on March 30, 1969, just before the stream was subjected to
wastes by direct discharge. During April the minimum was 14,500 col/100 ml.
In May the minimum was 31,000 col/100 ml and in June 300,000 col/100 ml.
During May and June the minimum values occurred after flushing with runoff.
TOG values were larger than BOD5 values during the early part of the
study but the reverse was true during the latter part. This indicates
that wastes in the stream during the early part of the study were more
oxidized than those in the latter phase. During the early phase, a
larger portion of the wastes was from the drylots.
After the stream had become heavily loaded with wastes, TOC
exceeded the BOD,, for short periods during peak rate of runoff from
61
-------
0.8 r-
0.6
to
en
N
co -O
0.2
O MAX. DAILY TEMP.
A MIN. " "
FLOW
'— RAIN
TIME OF
PEAK FLOW
I
10
20
30
TIME OF
WASH-DOWN
EVENT
40 0
TIME, HOURS
O
100
80
TIME OF
PEAK FLOW
TIME OF
PEAK FLOW
60 H
M
40
20
40
Figure 19. Hydrological Data, Site D, May 25-26 and June 12-13, 1969.
-------
•a
o
o
81-
w
H
O
u
1*4
O
o
rH
ON
TIME OF
PEAK FLOW
TIME OF
.WASH-DOWN
EVENT
I
TIME OF
PEAK FLOW
r
TIME OF
PEAK FLOW
10
20
30
40 0
TIME, HOURS
10
20
30
40
Figure 20. Fecal Coliform, Site D, May 25-26 and June 12-13, 1969.
-------
500
400
300
200
Q O
O O
PP H
100
0
TIME
r OF
I PEAK FLOW
1
1!
M 11
TOC
BOD
TIME OF
WASH-DOWN
EVENT
TIME OF
PEAK FLOW
J
•X
r TIME OF
IPEAK FLOW
I I
10
20
30
40 ' On
TIME, HOURS
10
20
30
40
Figure 21.
BOD5 and TOC, Site D, May 25~26 and June 12-13, 1969.
-------
80 i—
KJELDAHL
AMMONIA
W
tO
a
w
e>
§
H
60
40
20
TIME OF
PEAK FLOW
III
I Ml
i n
TIME OF
SH-DO
EVENT
\ WASH-DOWN , V
10
20
30
\
V
TIME OF PEAK FLOW
t
TIME OF PEAK FLOW
40 ' 0
TIME, HOURS
10
20
30
40
Figure 22. Kjeldahl and Ammonia Nitrogen, Site D, May 25-26 and June 12-13, 1969.
-------
100 r-
80
o
CH
w
cfl
60
00
B
W
H
CO
s
PM
40
IM
'
20
TIME OF
PEAK FLOW
II
\\
TlME OF
|| , WASH-DOWN
EVENT
i d I
iTIME \OFj
[PEAK
IFLOW
10
20
30
40 ' 0
TIME, HOURS
10
20
30
40
Figure 23. Phosphate, Site D, May 25-26 and June 12-13, 19691
-------
0.6
to
03
« 0.4
O
is
0.2
TIME OF
WASH-DOWN
EVENT \
I
I
TIME OF
PEAK FLOW
10
20
30
40 " 0
TIME, HOURS
10
20
30
40
Figure 24. Nitrate Nitrogen, Site D, May 25-26 and June 12-13, 1969.
-------
greater runoff events. This suggests that contributions of waste from
land origin, e.g., the drylot operation, or from sources of more stable
and/or less degradable materials were entering the stream. During the
first part of the runoff event and shortly after peak runoff, the 6005
exceeded the TOG. This indicates that a major portion of the wastes in-
the stream was from the direct discharge sources.
Higher values of N03 were generally obtained during periods of sur-
face runoff than during wash-down events. This suggests that the waste
in the runoff waters had undergone considerable aerobic decomposition and
further substantiates the explanation of the reversal in the BOV,./TOC
ratio during greater runoff events.
The wastes washed directly into the stream from the confinement
growing areas not only caused high pollution indice values during the
wash-down events but also caused high values during base-flow conditions.
Furthermore, the large amounts of wastes deposited on the bed of the
stream between runoff events caused the pollution indices to rise
rapidly at the beginning of runoff events, a first flush effect. During
light runoff events, the pollution indices normally peaked at peak flow
and showed a corresponding decrease with runoff rate. However, during
larger runoff events, the pollution indices peaked before peak runoff
and showed a decrease during high runoff rates followed by a second peak
as runoff decreased. After this second peak, they showed a corresponding
decrease with runoff rate to values below base-flow contents. As runoff
stopped, the indices returned to the base-flow values. This phenomenon
is illustrated by Figures 19-24.
Results of calculations of the amount of pollution/day/H carried by
the stream at site D are shown in Table 7. The value for FC/day/H was
obtained by multiplying the mean of all FC measurements made on the
samples from stream D by %. The value for TC/FC was obtained from all
paired TC and FC measurements made on the samples from stream D. The
other results were obtained in a similar manner.
Land Runoff Studies
Examples of runoff water quality data and corresponding hydrological
data for sites E, F, and K are presented graphically in Figures 25-29.
Wastes spread on site E Were evaluated by dividing the product of
the average H and the time period of waste production in the confine-
ment growing operation into the amount of waste produced to obtain the
amount of waste per H per unit time. This value was then divided into
the appropriate rate of spreading selected from Figure 2 to obtain H
values for wastes spread. These values were added to the H values for
hogs on pasture and the sum was rounded off to give the values reported
for site E.
Site F, the control area, was free of swine and all other domestic
animal wastes. Thus, H is reported as 0. H values reported for site K
were obtained from reports by the swine producer of the total weight of
hogs in the drylots.
68
-------
Table 7. Summary of the effect of discharging swine wastes directly
into stream D
Variable
! i I i ntaeaasasa • •' • . i s=ss=s
H
Ef s
al/day/H
FC
TC
col/100 ml
col/ day /H
col/ 100 ml
col/day /H
TC/FC
BOD |mg/1
|lb/day/H
TOC
mg/1
Ib/day/H
BOD5/TOC
N
mg/1
Ib/day/H
TN/NH3-N
P04
mg/1
Ib/day/H
TP/OP
Max
300
6.3
36 x 106
24 x 106
520
380
72
60
Mean
====i===
245
aioa
290
2.7 x 106
28.5 x 109
4.0 x 106
44 x 109
1.5
72.7
.167
52.9
.122
1.20
13.0
.030
1.40
6.9
.016
1.5
Min
100
.057
150
150 x 103
6.8
10.5
2.0
.5
Fresh waste
9b
8.9 x 10
.32C
-048C
.035°
V
bGieldrich (1966).
CMuehling (1969).
Linear correlation coefficients for H with most of the pollution
indices showed significant correlations for data groups E and K. At
site K, H was significantly correlated at the 0.01 level with TC, BOD ,
TOC, TN, and NO.,. The correlation coefficient between H and TC should
69
-------
s
O
CO
P4
o
P4
MAX. DAILY
TEMP.
TIME OF PEAK FLOW
MLN. BAIL'
TEMP.
40 » 0
TIME, HOURS
10
TIME OF PEAK FLOW
lOO
80
60
H
M
W
40
20
20
30
Figure 25. Hydrologlcal Data, Site E, March 18-19 and May 19-20, 1969.
-------
SI-
TOTAL COLIFORM
FECAL COLIFORM
o
w
M
§
h4
8
fn
O
o 4
,-1
§
a
w
B
<
PQ
l\
TIME OF
PEAK FLOW
TIME OF
PEAK FLOW
\
10
20
30
40 0
TIME, HOURS
10
20
30
40
Figure 26. Bacteria, Site E, March 18-19 and May 19-20, 1969.
-------
80 r-
60
00 M
6 6
n u
o o
« H
40
to
2C~
TIME OF
PEAK FLOW
BOD
TOC
I
TIME OF
PEAK FLOW
10
2Q
30
40 Q
TIME, HOURS
1Q
2Q
30
40
Figure 27. BOD5 and TOC, Site E, March 18-19 and May 19-20, 1969.
-------
8_
en
Kl
oo
W
o
§
H
25 2 -
KJELDAHL
AMMONIA
NITRATE
10
20
40 0
TIME, HOURS
10
Figure 28. Kjeldahl, Ammonia and Nitrate Nitrogen, Site E, March 18-19 and May 19-20, 1969.
-------
40 ' 0
TIME, HOURS
Figure 29. Phosphate, Site E, March 18-19 and May 19-20, 1969.
-------
not be regarded as meaningful since only 9 TC values were included and
8 of these were obtained for a constant H value. The poor correlations
within most data subsets was partially due to the small variations in the
H value throughout the periods covered by these data groups. When all
data for a given site were combined, the correlations between H and the
other indices were evident because H varied during the periods covered.
Even so, the correlation coefficients for data group E were rather low.
This was also the case when data from sites E and K were combined as
data group EK. These latter correlations suggest that the H values reported
for site E may have been poor. Since better correlations were obtained
from site K, solely a drylot operation, than from site E, the problem was
probably associated with H values. If this were the case, then the prob-
lem probably arose from failure to consider the decomposition of wastes
in the pits before it was spread on the land and the differences in fate
of the spread wastes compared to those voided in drylots. Observations
of the spread wastes revealed that the solids were well dispersed. Thus,
there was greater opportunity for attachment to the soil than when
solids were excreted in lumps on the drylots. Also, the wastes were
spread on a heavy growth of grass and a considerable portion lodged
above the ground surface. This may have allowed rapid decay with
decreased movement to the stream.
Another probably cause for the low correlations with H is that only
a small part of the total watershed was subjected to swine wastes. As
suggested by the study at site F, natural pollution from areas free of
domestic animal wastes may account for a substantial portion of the total
stream pollution associated with land on which animal wastes are disposed.
Pollution indice values for the raw wastes spread on site E varied
widely depending on the amount of water involved and the age of the wastes.
The data obtained from analyzing 18 samples collected from the manure
wagon used to spread the wastes gave the following results:
Indice
TC
FC
BOD
TOG
TS
VS
TN
TP-P04
OP-PO,
Max
130xlO€
87x10*
62,000
43,000
344,000
264,000
12,700
8,500
7,000
Mean
27xlO£
13«10£
25,000
14,000
220,000
180,000
4,100
2,700
1,500
Min
6xl06
4xl06
4,600
3,300
5,700
4,600
320
115
75
Units
col/100 ml
col/100 ml
mg/1
mg/1
mg/1
mg/1
mg/1
mg/1
mg/1
Variations in indice values between samples were expected because of the
differences in the amount of dilution and sampling error. When compared
to wastes entering a lagoon, the much higher indice values obtained for
these wastes were due to differences in dilution.
75
-------
Hydrologlcal Data. Generally, minimum temperature, Tn, was a little
better correlated with the pollution indices for land runoff samples
than was maximum temperature, TX. The correlations of TX and Tn with
pollution indices were low for the data groups containing only data from
sites E and K. The study period for site E ran from the beginning of
1969 through August 1969 and gave temperature variations large enough to
point up correlations if they had existed. The study period for site K
was limited to the spring and summer of 1969; and thus, the temperature
variations were small and may account partially for the poor correlations.
Variations in the hydrographs for the streams located at the land
runoff sites were generally associated with rainfall- runoff events.
Small daily variations were discernible only at site K. Because the
rainfall and flow data collected during the study were insufficient to
allow development of stream flow prediction equations, measured stream-
flow values were used in the regression analyses that were performed to
obtain predictions of pollution indice values in the streams.
Linear correlation coefficients between q^ and pollution indices
were slightly better than for either qt or qo for data group E. How-
ever, subsets of data group F often showed that the correlations were
better for qt and q0. At sites E and K, the pollution indices were more
closely related to qt and qo. Since the swine wastes reaching the streams
were hypothesized to be associated with overland flow, q was
an independent variable for the regression analyses.
selected as
Since sampling was commonly associated with rainfall-runoff periods,
the qt values for sites E, F, and K obtained by averaging the flow rates
at the time of sampling were greater than the average flow rates. Closer
estimates of the average flow rates were given by the qjj values. The q^
values at times of sampling were higher than the average q^ values
because q^ also increased during rainfall-runoff periods. The results
presented in Table 8 were obtained using q^. The flow rates represent
water yields of 9.9, 12.6, and 8.2 in./ac/yr for sites E, F, and K,
respectively. Larger maximum flow rates were obtained at sites E and K
than at site F even though the drainage areas at E and K were smaller.
A major factor causing this was the differences in slopes and surface
cultures. Basins E and K were farmed more intensely and row crops were
grown on poorer classes of land than was true for watershed F. Thus, higher
rates of runoff, and erosion, resulted.
Table 8. Rates of streamflow from the land runoff sites
GROUP
(Site)
E
F
K
max q tn min q
to t
4.9
2.1
5.4
cfs
.040
.109
.047
.021
.025
.001
q-
106ft3/yr
1.26
3.45
147
gal/ day /ac
740
940
590
gal/day /H
51
-
154
76
-------
B_acteria Indices. Bacteria counts rose considerably during runoff
events at sites E, F, and K. Correlations of FC and TC with qt and q
were good for several of the subsets of data groups E, F, and K. The
correlations were also good for data group K; but when all data for
site E were examined, there was almost no correlation between these
variables. The same was true for the data from site K. Over the long
period, then, the relations between flow and bacteria levels were over-
shadowed by other more important factors.
Both TC and FC counts increased for all flow conditions at sites E
and F as the study progressed. This was probably caused by temperature
differences, because the temperature rose during the course of the
study and provided a warmer and more favorable growing environment. The
significant positive, although low, correlation coefficients for data
groups E and F suggest that the increase in temperature during the study
period at least partially caused the observed increases.
In general, correlations between FC and fC and the other pollution
indices were good for sites E and K. The poorer correlations at site F
were related to low indice values. For example, the NH -N value for
group F was only 0.3 mg/1. J
Results of calculations of the bacteria carried by the streams are
shown in Table 9. The value for FC/day/H for data group E was obtained
by multiplying the mean of all FC measurements made on samples from site E
by q" for data group E from Table 8. The value for TC/FC for data
group E was obtained from all paired TC and FC measurements made on
samples from stream E. The other results were obtained in a similar
manner. Data group F2 (control site subset) is included in Table 9 and
the following discussion because it covers the same period of data
collection as data group K.
Table 9. Bacteria in streamflow from land runoff sites
GROUP
(Site)
E
F
K
F2
FC
col/ 100 mo
max
8.3xl06
.8xl06
11.4xl06
. 8xl06
mean
189xl03
lOxlO3
365xl03
21xl03
nun
2
1
1450
19
10 col/day/ac
5,120
330
8,200
750
106 col/day /H
358
_
2,040
-
TC/FC
2.1
3.9
2.4
3.8
The TC/FC ratio was generally lower on sites E and K than on site F.
This followed the expected pattern because the source of FC is primarily
from animal wastes whereas TC may arise from other sources as well; e.g.,
soil. Thus, the lower TC/FC ratios at sites E and K reflect a larger
portion of animal wastes in the stream waters. The TC/FC ratio was also
lower at all sites during periods of rainfall-runoff than during periods
of no surface runoff. This suggests that during rainfall-runoff periods,
some animal wastes were washed into streams, including the control
stream (wildlife).
77
-------
BOD5 an<^ T^C. Excellent linear correlations generally existed
between BOD^" and TOC at sites E and K whereas the correlation coefficients
at site F were generally lower. Essentially, all correlations were
positive and significant at the 0.01 level. Better correlations resulted
when the stream samples were more polluted. When the streams contained
low amounts of biodegradable wastes, the BOD5 values were questionable
because the amount of oxygen used in the test was generally below the
recommended amount of 2 mg/1.
BOD5 and TOC usually showed good correlation with the other pollution
indices for the land runoff data groups. Thus, both BOD5 and TOC pro-
vided good measurements of the pollution loads on the streams.
Results of calculations of BOD5 and TOC carried by the streams are
shown in Table 10. The value for BOD5/day/H for data group E was
obtained by multiplying the mean of all BOD5 measurements made on samples
from site E by q~ for data group E from Table 8. The value for BOD5/TOC
for data groups E was obtained from all paired BODc and TOC measurements
made on samples from stream E. The other results were obtained in a
similar manner.
TOC values were larger than BOD^ values for the samples collected
from the three land runoff sites indicating that the wastes which reached
the streams were relatively well oxidized. The higher ratios for sites E
and K suggest that wastes entering the streams from the swine production
areas were generally less degraded biologically than natural wastes from
the control site.
N i tro gen» Ammonia nitrogen was a higher percentage of the TN on
the waste watershed than on the control. The maximum TN value of the
317 land runoff samples analyzed was 15.6 mg/1. This sample was collected
from site F. The average TN content of all 317 samples was 1.2 mg/1.
The TN concentrations under base flow conditions were 0.1, 0.2, and 0,5
mg/1 at sites E, F, and K, respectively.
The maximum NH^-N value of the 186 samples analyzed was 6.8 mg/1.
As with TN this sample was also collected from site F. The maximum
values at sites E and K were 6.0 and 1.7 mg/1, respectively. The average
of all 186 samples was 0.4 mg/1. The means of the samples for sites E,
F, K, and F2 were 0.6, 0.3, 0.4, and 0.4 mg/1, respectively.
The NO.,-N content was about 0.01 mg/1 under base flow conditions
at sites F and K and was near 2.6 mg/1 at site E. The nitrate concen-
trations on watershed E were generally inversely related to surface
runoff. This was particularly true during the winter and early spring
period. Later in the study, the NOg content showed both increases and
decreases with increased flow rates. The increases were associated with
high rates of surface runoff. Essentially, all data groups for sties F
and K data showed positive correlation coefficients.
Results of calculations of the nitrogen carried by the stream are
shown in Table 11. The value for N/day/H for data group E was obtained
by multiplying the mean for all TN measurements made on samples from
78
-------
Table 10. BOD5 and TOG in streamflow from the land runoff sites
GROUP
(Site)
E
F
K
F2
BOD5
mg/1
max
70.0
16.8
35.0
16.8
mean
4.7
2.0
6.4
2.7
min
.1
.1
.9
.1
Ib/day/ac
.028
.015
.031
.021
Ib/day/H
.0020
-
.0079
-
TOG
mg/1
max
85.0
380.0
140.0
380.0
mean
9.6
14.6
17.6
22.7
min
1.0
3.0
5.8
5.0
Ib/day/ac
.057
.111
.087
.177
Ib/day/H
.0040
-
.0216
-
BOD5/TOC
.51
.17
.36
.18
Table 11. Nitrogen in streamflow from the land runoff sites
GROUP
(Site)
E
F
K
F2
TN
mg/1
max
12.0
15.6
7.7
15.6
mean
1.1
1.2
1.4
1.6
min
.1
.1
.3
.1
Ib/day/ac
.0068
.0094
.0069
.0125
Ib/day/H
.00047
-
.00180
-
NOo-JN
mg/1
max
9.99
.68
2.00
.50
mean
2.58
.20
.30
.17
min
.55
.01
.01
.01
Ib/day/ac
.0159
.0016
.0015
.0013
Ib/day/H
.00110
-
.00039
-
TN/NH -N
2.34
4.00
3.75
3.00
-------
site E by q for data group E from Table 8. The value for TN/NI^-N for
data group E was obtained from all paired TN and NBU measurements made
on samples from stream E. The other results were obtained in a similar
manner.
Phospha te. TP and OP concentrations followed the same hydrograph
pattern as the other indices, except NC>3 at site E, of increasing and
decreasing with flow rates. Base flow TP contents were between 0.25 and
1 mg/1. Maximum TP values occurred for the samples that contained the
largest amounts of soil. This indicates that the phosphorus entering
the streams was closely related to erosion.
Results of calculations of the phosphorus carried by the streams
are shown in Table 12. The value for PO^/day/H for data group E was
obtained by multiplying the mean of all TP measurements made on the
samples from stream E by the q" for data group E from Table 8. The
value for TP/OP for data group A was obtained from all paired TP and
OP measurements made on the samples from stream E. The other results
were obtained in a similar manner. Phosphorus entering the streams was
mainly OP and thus was immediately available to plants.
Table 12. Phosphate in streamflow from the land runoff sites
GROUP
(Site)
E
F
K
F2
P°4
mg/1
max
17.0
4.0
13.0
4.0
mean
1.2
.2
1.9
.3
min
.1
.1
.1
.1
Ib/day/ac
.0074
.0016
.0094
.0024
Ib/day/H
.00051
-
.00234
-
TP/OP
1.16
1.00
1.15
1.00
Discussion. There was no evidence of pollution from swine waste
spreading at site E during periods of dry-weather flow except for
nitrate. While negligible, there was some indication of pollution from
the drylot operation at site K which is attributed to erosion. During
rainfall-runoff events, however, pollution indices from runoff at sites
E, F, and K rose to unexpectedly high levels in response to increasing
rates of flow. With the possible exception of bacterial densities,
pollution through surface runoff from areas utilized for land disposal
of swine wastes was not markedly higher than from the control watershed
which was entirely free of farm animals.
The higher than expected N03 content at site E suggests the possi-
bility of groundwater pollution and discharge in the base flow of the
stream. Since the nitrate concentration at site K was not significantly
different from that at site F, the nitrate in the stream at site E is
80
-------
assumed to be due to the wastes spread on the watershed and not the
drylot source. The 135,000 gal of raw wastes spread during the study
period at site E was applied to only five acres of the forty-acre water-
shed. If this same rate, 27,000,gal/acre, had been applied to the total
watershed, the resulting N03-K content in the groundwater and base flow
would have been 18 mg/1. This value is 1.8 times the maximum value
specified by the Public Health Service for nitrate nitrogen in drinking
water. The application period for the wastes at site E was about six
months. Thus, a yearly application rate of 27,000 gal/acre over the
entire watershed would maintain the NO- content below the acceptable
drinking water standard. At this rate the disposal area required would
be 1 ac/100 H when the entire watershed is used. In addition, this
lower rate of application would enable plants to use more of the nitrogen
and an even smaller amount would then enter the stream via groundwater.
The effects of animal wastes on the study sites were often over-
shadowed by watershed factors, particularly the slopes and surface
cultures. These were important in determining the amount of erosion and,
thus, the amount of pollution reaching the stream. For example, the
PO^ loads on stream K when compared to stream E suggest that five times
as much PO^ per animal reached the stream at site K as at site E.
The lower bacteria densities from the control watershed as compared
to the generally higher values from the waste watersheds suggests the
possibility of some stream contamination from runoff due to land disposal
of animal wastes. The BOD/TOC ratio for the watersheds tends to confirm
this assumption. Differences in the calculated numbers of colonies per
day per H between sites E and K suggest that the occasional higher
numbers at site K were due to the swine having direct access to the stream.
The data obtained from site E during March 18-19 and May 19-20, 1969,
(Figures 25-29) are representative of data obtained from the land runoff
sites. All pollution indices except N03 at site E increased substantially
with surface runoff and paralleled the hydrographs with extended drag-
out on cessation of surface runoff. Changes in magnitude of the indices
were generally greater for higher rates of surface runoff. This is
exemplified by comparing the values obtained during the March and May
periods except for bacteria counts. Base-flow bacteria counts were also
higher during the second period than during the first. This followed
the general trend of higher counts during warmer and more favorable
temperature periods. Another reason for the higher bacteria values
during the May period is the difference in surface culture. The entire
basin at site E was well protected with vegetation during the March
period. Land preparation had begun during the May period and some areas
were subject to easy erosion. A less likely reason may have been that
wastes were spread immediately preceding and during the rainstorm in
the May period whereas the most recent waste spreading in the March
period was five days preceding surface runoff.
The data shown in Table 13 from sites E and F illustrate the effect
of runoff rates on pollutional discharges. The period of surface runoff
was about 10 hours at site E and 25 hours at site F during the period
March 18-25. The peak-flow rate was 90 times the base-flow rate at
site E and only 20 times at site F.
81
-------
Table 13. Response of stream pollution to flow rate - March 18-25, 1969
Variable
q , cfs
Ht*
FC, col/ 100 ml
TC, col/100 ml
BOD5, mg/1
TOG, mg/1
TN, mg/1
NH3-N, mg/1
N03~N, mg/1
TP-P04, mg/1
OP-P04, mg/1
_Haste Watershed (site E)
Base Flow
.04
45
120
.5
1.0
.1
.1
2.65
.1
.1
Peak Flow
3.70
400,000
500,000
39.0
63.0
12.1
5.1
.55
9.0
8.2
Control Watershed (site F)
Base Flow
.114
19
20
1.5
5.0
.1
.1
.15
.1
.1
Peak Flow
2.30
6,500
8,667
4.6
21.5
3.0
.2
.46
.7
.4
Dairy Waste Study
Hydrological Data
of Tx and Tn with the pollution indices were low at
The correlations
the dairy sites although the temperature variations during the study were
large enough to point up correlations if they had existed. Also, the
linear correlation coefficients between qt, q^, and q with the pollution
indices showed that the flow rates were poor measures of pollution in
the streams. qt and qo were usually better correlated with the pollution
indices than was qj, and the rate of streamflow was a better measure of
the degree of pollution in stream H than in stream J.
The flow data presented in Table 14 were obtained by using the
average of the base-flow rates in the stream at the times of sampling.
The flow rates represent yields of 7.8 and 30.7 in/ac/yr from sites H
and J, respectively. Two major factors attributed to the differences
in yields. The surface culture on watershed J was poor due to the dense
concentrations of animals and the building sites, while vegetation
generally covered the entire watershed at site H. This resulted in more
surface runoff at site J than at site H. Although watershed J was
smaller than watershed H, the amount of wash water dumped into the
stream was greater at site J due to the larger number of animals and the
use of more wash water per animal necessitated by the poorer job of
solids removal before wash-down.
Pollution Indices
All pollution indices were high in the streams receiving dairy
wastes. Fecal coliform counts were exceptionally high. Stream J was
more polluted than stream H. Visual observations of the amounts of
solid animal wastes carried by the streams showed that stream H was
relatively free of solids while stream J was more heavily loaded than
stream D (which received direct discharges of swine wastes).
82
-------
Correlations between pollution indices were quite high. BODr and TOC
generally showed sufficient correlation with the other indices to suggest
either as a good measure of the degree of stream pollution and the amount
of dairy wastes which entered the stream. COD and N03 were often poorly
correlated with the other indices which suggests they were the poorer
measures of stream quality.
Results of calculations of the amount of pollution carried by the
dairy streams are shown in Table 14. The values of FC/day/cow at site H
were obtained by multiplying the mean of all FC measurements made on the
samples from stream H by q^ for stream H. BOD5/COD ratios at site H
were obtained from all paired BODs and COD measurements made on the
samples from stream H. The other results were developed in a similar
manner. The data in Table 14 show that the FC was 875 percent at site H
and 1340 percent at site J of the average production rates from fresh
cow wastes reported by Gieldrich (1966). The BOD5 was 6 percent and
23 percent, the COD was 11 percent and 16 percent, and the N (TN plus
N03-N) was 10 percent and 10 percent for sites H and J, respectively,
of the average production rates for fresh dairy cow wastes reported by
Witzel, £t al_ (1966).
Discussion
The exceptionally high values of FC carried by the dairy streams
suggest an overgrowth of bacteria occurred in the streams above the
sampling points. This overgrowth was limited to warmer weather. During
the winter period, the FC counts in the streams were low which suggests
that most FC bacteria died off before reaching the sampling station. An
overgrowth of FC was also observed in stream D which was subjected to
fresh swine wastes by direct dumping.
Even though the pollution indices other than FC indicated that only
fractions of the dairy wastes reached the streams, the amounts were still
sufficient to cause gross pollution conditions. For example, the mean
BOD5 shown in Table 14 for stream H is 17.1 mg/1 and for stream J is
167 mg/1.
The amounts of wastes carried by the streams varied erratically
with time and were poorly related with flow rates and temperature. The
wastes washed directly from the milking parlors overshadowed the wastes
reaching the streams from the surrounding pastures. However, the streams
were generally more polluted during the summer than during the winter and
pollution increased with surface runoff.
Poultry Waste Study
Hydroloeical Data
The correlations of Tx and Tn with the pollution indices were
usually low at the poultry waste sites although some significant correla-
tion coefficients were obtained at site P where large temperature
variations occurred during the study period covered. Even at site P,
however, the correlations were too inconsistent to suggest a relationship
between temperature and the quality of the stream water.
83
-------
Table 14. Summary of surface water pollution from the dairy sites.
max
Area, ac
Cows
q
FC
COD
cfs
106 ft /yr
gal/day/ac
gal/day/cow
col/100 ml 78 x 106
106 col/day/ac
106 col/day/cow
mg/1 1525
Ib/day/ac
Ib/day/cow
mg/1 350
BOD,.
J
TOG
Ib/day/ac
Ib/day/cow
mg/1 520
Ib/day/ac
Ib/day/cow
BOD5/COD
BOD5/TOC
TOC/COD
H
mean
70
75
.063
1.99
580
543
2.3 x 106
51,100
47,200
139
.675
.630
17.1
.083
.078
28.0
.136
.127
.18
.59
.33
Site
J
min max mean
15
155
.053
1.67
2280
222
1 215 x 106 8.6 x 106
725,000
72,300
2.0 2930 504
9.58
.935
.1 1430 167
3.08
.308
3.0 1790 182
3.36
.337
.28
.73
.32
min
28 x 103
5400a
32.5
h
5.78b
10.0
b
1.32
24.0
-------
Table 14. (Continued)
Variable
max
Site
H
mean
mm max
mean
Fresh Wastes
mm
TN
mg/1 40.0
Ib/day/ac
Ib/day/cow
7.1
.0345
.0322
.7 90.8 18.6
.3540
,0335
1.5
.37b
1.41
3.37
CO
Ui
N03-N
PO,
mg/1
Ib/day/ac
Ib/day/cow
|mg/l
llb/day/ac
Ib/day/cow
5.85
30.5
.75
.0036
.0034
4.6
.0224
.0208
.05 10.0
.5 62.0
2.19
.0417
.0041
18.3
.3485
.0339
.02
.3
aGieldrich (1966).
bWitzel et al. (1966).
-------
The rates of flow, qt, qb and qo, were poor measures of the degree
of pollution from sites P and X. The base flow at site X was constant
at zero since the station was located in a meadow strip and thus no
correlation could be expected between q^ and the pollution indices for
the site. The negative correlation coefficients for qt and qo with the
pollution indices obtained for data group X2, the data group for the
runoff period shortly after the poultry wastes were spread, indicates
that the runoff was more polluted at low rates of flow. Higher flow
rates resulted in lower concentrations of wastes although the amount of
wastes which reached the station was greater on a unit time basis.
The flow data presented in Table 15 for site P were obtained by
using the average base flow rate in the stream at the time of sampling
as the average water yield from the watershed. The flow rate repre-
sents a yield of 7.4 in/ac/yr. The flow rate at site X averaged 0.006
cfs during the period of study. This value was used to obtain the
results in Table 15 for site X and represents a yield of 10.4 in/ac/yr.
Pollution Indices
The data in Table 15 pertaining to the number of hens contributing
wastes at the two sites were selected on the following basis: Site P
received wastes from a total of 10,500 hens per year in three equal
applications. Since the wastes were spread on regular intervals of
about four months, the number of hens was taken as 10,500. At site X
only one application of wastes was made. The wastes were equivalent to
the yearly wastes from about 3200 hens. The number of hens contributing
wastes was taken as 9600 by assuming that a total of three such additions
of wastes could be made per year with similar results of runoff amount
and quality.
The pollution indices were often significantly correlated with one
another. Exceptions were normally associated with low index values
which suggests the tests were too inaccurate to point up the correlation
on the less-polluted samples.
In view of the amount of natural pollution measured in stream F,
the control stream for the swine waste study, the pollution observed in
stream P was most likely from natural sources. No change in the quality
of the stream was discernible following the two periods of waste spread-
ing as compared to the quality before spreading. The pollution indices
followed the same trends observed in stream F of increasing surface run-
off. Also, FC counts increased with warmer temperatures.
The pollution indices at site X before the poultry wastes were
spread were quite high. All index values increased after waste spreading,
reflecting considerable poultry wastes in the runoff. For example, the
average BOD5 increased from 9.8 mg/1 to 370 mg/1. Pollution indices for
runoff two months after the wastes had been spread showed that some
poultry wastes were still being removed from the field. For example, the
average BOD5 was 22.5.
Results of calculations of the amount of pollution carried by the
poultry streams are shown in Table 15. The value of FC/day/hen at site P
86
-------
Table 15. Summary of surface water pollution from the poultry sites.
Variable
max.
Area, ac
Hens
cf s
FC
COD
BOD5
TOC
106 ft /yr
gal/day/ac
gal/day/hen
col/100 ml 160000
106 col/day/ac
col/day/hen
mg/1 535
Ib/day/ac
Ib/day/hen
tng/1 22
Ib/day/ac
Ib/day/hen
mg/1 41
Ib/day/ac
Ib/day/hen
BOD /COD
•J
BOD /TOC
*J
TOC/COD
TN
mg/1 24.6
Ib/day/ac
Ib/day/hen
TN/NH--N
3
NO -N
PO
mg/1 9.00
Ib/day/ac
Ib/day/hen
mg/1 7.9
Ib/day/ac
Ib/day/hen
SITE
P
mean min max.
65
10500
.055
1.73
547
3.4
9600 1 1050
190
1.2 x 10°
63.2 1 1720
.285
0.0014
5.2 .1 1070
.023
0.0001
14.2 4.0 1108
.063
0.0004
.086
.298
.313
1.9 -3 270
.009
0,00005
3.17
1.55 -01 4.85
.007
0.00004
1.2 -1
f\r\C.
.005
0.00003
X
mean
5
9600
.006
.19
111
.4
235
6.9
3600
600
3.90
.0020
264
1.66
.0009
322
2H9
• \)£*
.0011
.422
.844
.507
60
.390
.0002
HI *» —
2.00
.013
0
"""
mm
50
33
7.5
28
.1
.05
87
-------
was obtained by multiplying the mean of all FC measurements made on the
samples from stream P by q" for stream P. The value for BOD5/COD at
site P was obtained from all paired BODr and COD measurements made on
the samples from stream P. The other results were obtained in a similar
manner.
Although the results in Table 15 include contributions from natural
sources, the amounts of pollution per day per hen are less than 1/10 of
the average production rates from fresh poultry wastes. For example, the
BOD5 carried by stream P, 0.00014 Ib/day/hen, accounted for only 0.6
percent of the BODc production by the hens based on an average production
rate of 0.024 Ib/day/hen.
Discussion
The data obtained from study of the two poultry waste disposal
sites suggest that land spreading can be effective in controlling stream
pollution. However, the application of wastes at high rates on bare
soil and in drainage ways, as was the case at site X, results in consider-
able water pollution. This points to the need for criteria to guide the
selection and management of disposal areas. However, even in cases
where the site of disposal is poorly located, the amount of pollution
which reaches the stream is a very small proportion of the potential
pollution from poultry wastes.
Beef Waste Study
Hydrological Data
The correlations of TX and Tn with the pollution indices were low
almost without exception at the beef pasture site although sufficient
variations occurred in the temperature to point-up relationships if they
had existed. Correlations between flow rates and the pollution indices
were also poor. Nutrient contents in the stream were low and may have
accounted for the lack of correlation. FC counts were higher during warmer
weather. This dependence probably caused the lack of correlation between
FC and flow rates.
The flow data presented in Table 16 were obtained by averaging the
flow rates in stream Z for the period of study. In this case the
average base flow rate at the times of sampling, 0.007 cfs, was not a
good estimate of the average flow rate in contrast to what was true at
the other study sites. This difference was due to a lack of increase in
base flow at site Z when rainfall-runoff events occurred and was probably
associated with the geology of the watershed.
Pollution Indices
The pollution indices at site Z were often significantly correlated
with each other except when the values of the indices were low. As was
the ".ase. with the other land disposal sites, BOD5 and TOG showed the
best correlations with the other indices.
88
-------
Table 16. Summary of surface water pollution from the beef site.
Variable
Area, ac
Cows
q
FC
COD
BOD""
T(5(f
cfs
106ft /yr
gal/day/ac
gal/day/cow
col/100 ml
106 col/day/ac
10^ col/day/cow
mg/1
Ib/day/ac
Ib/day/cow
mg/1
Ib/day/ac
Ib/day/cow
mg/1
Ib/day/ac
Ib/day/cow
BO»5/COD
BOD5/TOC
TOG/COD
TN
mg/1
Ib/day/ac
Ib/day/cow
TN/NH3-N
N03-N
»;
mg/1
Ib/day/ac
Ib/day/cow
mg/1
Ib/day/ac
Ib/day/cow
Max Mean
25
35
.020
.63
517
370
460,000 30,700
580
430
695 149
.645
.460
50 9.8
.041
.0303
206 32.5
.136
.1000
.080
.330
.278
3.1 1.2
.0052
.0037
6.0
3.20 1.35
.0058
.0042
1.9 1.1
.0048
.0034
Min
1
8.0
.3
6
.3
.01
.4
-------
Results of calculations of the amounts of pollution carried by stream
Z are shown in Table 16. The value of FC/day/cow was obtained by multi-
plying the mean of all FC measurements made on the samples by the average
flow rate. The value for BODs/COD was obtained from all paired BOD5 and
COD measurements. The other results were obtained in a similar manner.
By using the same values for fresh waste production as were used for the
diary cows, the FC was 8 percent, the BOD5 2 percent, the COD 8 percent,
the P04 2 percent, and the N (TN plus N03~N) 2 percent of the average
production rates for fresh wastes.
Discussion
The stream draining the beef pasture site was more polluted than
would be expected from natural sources indicating that some wastes from
the animals did reach the stream. However, the amounts of pollution
atrributable to the beef animals was extremely low compared to the wastes
produced. Most of the pollution which could be attributed to the animals
was likely associated with wastes voided in the stream and the drainage
paths leading to the stream.
90
-------
SUMMARY
The purpose of this study was to investigate the actual and poten-
tial importance of animal wastes in agricultural land runoff with respect
to water quality management in the Southeast. It included:
1. the measurement of rainfall, runoff, bacterial densities,
organic materials and nutrients carried to streams from
swine, dairy, poultry, and beef growing areas;
2. the development of data and relationships for estimating
the effects of animal waste disposal under similar condi-
tions at other locations;
3. the assessment of present animal waste disposal practices;
4. recommendations for corrective action; and
5. identification of researchable problem areas.
Twelve sites were studied. These included six swine, two dairy,
two poultry, one beef, and one control watershed. Waste disposal prac-
tices at each of the sites were as follows:
Swine
Dairy
Poultry
Beef
Control watershed
Anaerobic lagoons 3 sites
Land spreading 2 sites
Direct discharge 1 site
Partial separation of
solids with land spread-
ing and direct discharge
of remaining wastes 2 sites
Land spreading 2 sites
Land spreading (grazing) 1 site
1 site
Sites are typical of present animal growing and waste handling practices
in the Piedmont Region of North Carolina.
Data were collected from each production site to ascertain the
amount and strength of wastes produced and eventually discharged to
receiving streams under varying conditions of rainfall, runoff, and
temperature. Time of flow-through measurements were made for the
lagoons. Attempts to determine the time of flow from land-runoff areas
used for waste disposal were unsuccessful.
91
-------
More than 1500 samples were collected in iced containers by auto-
matic samplers developed for the study. Analytical techniques commonly
used to determine the quality of waters and waste-waters were not all
directly applicable to the characterization of animal wastes, waste
lagoon effluents, and highly polluted stream samples. Interferences were
attributed to solids, dilution requirements, and toxic substances common
to these wastes. The biochemical oxygen demand varied directly with
dilution at BOB5 levels exceeding 60 mg/1. While unsatisfactory for
raw wastes and lagoon effluents, BOD was generally applicable to stream
samples. Difficulties encountered in automated nutrient analyses due to
high solids content were corrected by modifying the analytical techniques.
Sample preservation and storage for subsequent nutrient analyses using
freezing proved satisfactory but not ideal.
The total organic carbon analysis was applicable to all samples and
provided a convenient, rapid, and dependable method for determining
the organic strength of wastes and waste waters sampled during the study.
Good correlati9ns existed between BOD5 and TOG for the land runoff
samples but not the concentrated wastes. Difficulties common to the
standard BOD analysis for water containing high concentrations of animal
wastes (lagoon liquor and raw wastes) and the variability of BOD/TOG
with the degree of treatment precluded development of a general BOD/TOG
relationship for those wastes. Nevertheless, conjunctive use of BOD and
TOG parameters was useful in characterizing all samples and was particu-
larly useful when toxic materials and/or other factors limited the appli-
cability of the BOD test. The degree of BOD/TOG variability was one
indication of waste consistency and the possible presence of toxic
materials. "She BOB/TOG ratio also provided an indication of the ease of
biodegradation and/or the degree of stabilization.
Fecal streptococci were lower than expected. The lower counts
were likely due to the KF medium used. The applicability of the aedlun
to waters containing animal wastes was not checked.
The simple regression analyses revealed relatively high correlations
between many of the pollution indices which can be usefully developed
as a tool in water quality management. Certain of these are summarized
from Table 17.
Table 17. Correlation of TOG with other pollution indices for land-
runoff studies
Site
E (swine)
H (dairy)
PX (poultry)
Z (beef)
Control
r
FC
0.437
0.838
0.491
0.895
0.406
BOD5
0.966
0.970
0.610
0.724
0.734
TH
0.875
0.708
0.926
,0.955
0.943
TP
0 . 899
0. 714
0. 730
0.620
0.883
92
-------
The following simple regression equations with TOC as the independ-
ent variable and BOD5 as the dependent variable demonstrate the types of
potentially useful relationships which can be developed for the assess-
ment of pollution under applicable conditions:
Land runoff from
Site
Swine wastes E
BOD5 = 0.79 TOC - 2.5
Dairy wastes H
BOD5 = 0.66 TOC - 2.0
Poultry wastes P & X
BOD5 =0.35 TOC
TOC is particularly promising for this purpose because the analysis is
simple, rapid and reliable.
Multiple regression analyses using atmospheric temperature, stream-
flow, and number of animals as independent variables and the principal
pollution indices (FC, TC, BOD5> TOC, TN, NH3, N03, TP, and OP) as
dependent variables did not produce estimating equations of general
applicability. Inconsistencies between data groups with high coeffi-
cients of multiple determination were experienced. These are attributed
to vegetative cover, antecedent moisture, topography and soil types,
time of waste application with respect to rainfall, distance of disposal
site from stream, duration and intensity of rainfall, and possibly other
factors. However, high R values for some of the land runoff data groups
show promise that equations of general applicability for predicting
stream pollution from land runoff sites can be developed. Estimating
equations for BODc and TOC from the control watershed are examples.
Normative values for these indices range between 1-2 and 5-10 mg/1,
respectively, in the absence of runoff. The regression equation from
data group F23 with an r2 of 0.604
BOD5 = 2.8 - (0.035)(Tn) +2.2 (qQ)
produces an estimate of less than 1 mg/1 at 80°F and a flow of 0.10 cfs.
The TOC regression equation for the same data group has (r2 = 0.498)
TOC = 5.6 + (0.024)(Tn) + (13.9)(qo)
estimates a TOC of 9 mg/1 under the same conditions. The range in flow
experienced during the study was 0.025 - 2.1 cfs. These values are
consistent with typical base-flow conditions for land runoff study sites.
93
-------
Very little information is available on the quality of water from
natural watersheds. The following data from control watershed F indi-
cates that this can be appreciable under conditions of rainfall and
land runoff and needs to be taken into consideration in water quality
management.
Table 18. Average amounts of stream pollution from control watershed
Maximum
Mean
Minimum
FC
col/100 ml
800,000
10,000
1
BOD
mg/1
16.8
2.0
0.1
TOG
mg/1
380
14.6
3.0
TN
mg/1
15.6
1.2
0.1
J?i
mg/1
4.0
0.2
0.1
The normative amounts of wastes reaching receiving streams from
typical animal-growing operations are summarized in Table 19.
Table 19. Average amounts of wastes in land runoff from representa-
tive animal-growing operations
Swine .
Direct discharge—' .
Anaerobic lagoons-
Land spreading-!'
Dairyi7
Beef^
3/
Poultry-
EC
10& col/day
/animal
30,000
910
630
60,000
430
0.6
BOD5 TOG N P04
Ib /day / aiiima 1
0.176
0.154
0.003
0.190
0.030
0.0005
0.128
0.156
0.007
0.230
0.100
0.0007
0.032
0.009
0.002
0.034
0.008
0.0002
0.017
0.005
0.001
0.027
0.003
0.00003
— Sampling station downstream from point of discharge with inter-
vening section of stream providing some stabilization and settling.
21
— Lagoon effluent.
3/
— Sampling station at base of watershed immediately below site.
94
-------
These data vividly illustrate the importance of animal wastes as a
source of pollution. Wastes from a herd of 1000 swine following treat-
ment by anaerobic lagoon is roughly equivalent to the raw sewage from a
community of 1000 population. Land spreading as a means of disposal
reduces the BODt of the raw waste by at least 99 percent as compared to
about 50 percent for the lagoons. The present practice of dairy waste
disposal produces considerably more pollution on a per-animal basis and
is also unsatisfactory. Pollution from beef pasture, while only 15 per-
cent of the discharge from dairies is unexpectedly high. Animal access
to stream is an important causative factor. Poultry waste disposal by
land spreading was found to be entirely satisfactory in one case and a
major source of pollution in the other because of unsatisfactory spreading
practice.
Conclusions and recommendations are presented on pages 2-5 of this
report.
95
-------
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the contribution of
the following persons to this project:
Mr = Lloyd P. Tyler, Director, N.C. Department of Water
and Air Resources Laboratory; Mr. Darwin Coburn, Director,
Division of Water Pollution Control; and Mr. Earle C. Hubbard,
Assistant Director, Department of Water and Air lesources, for
services and consultation associated with nutrient analyses
using the Department's auto-analyzer.
Mr. James F. Koon, Mr. Amos F. Biles, and numerous
part-time employees for their dedicated field work and
laboratory analyses.
Dr. Robert S. Sowell, Dr. Sun-fu Shih, and Dr. Larry A.
Nelson for assistance with the data analyses.
Dr. Frank J. Humenik for consultation and advice on
technical problems.
Mr. Harold Snyder and Mr. Donald Anderson of EPA for
their assistance in the direction of the grant and the
preparation of this manuscript.
Dr. Jackie W. D. Robbing, Research Associate
Dr. David H. Howells, Project Director
Dr. George J. Kriz, Associate Department Head
Department of Biological and Agricultural Engineering
School of Agriculture and Life Sciences
North Carolina State University at Raleigh
96
-------
LIST OF REFERENCES
Anon, 1970. Watershed Research in Western North Carolina, Tennessee
Valley Authority and North Carolina State University, June.
APHA. 1965. Standard Methods for the Examination of Water and Waste-
water. 12th ed. American Public Health Association. Inc. New
York.
Azad, H. S. and D. L. King. 1965. Evaluating the effect of industrial
wastes on lagoon biota. Proceeding of the 20th Industrial Waste
Conference. Purdue University Engineering Extension Services
No. 118, p. 410-422.
Earth, E. F. , M. B. Ettinger, B. V. Salotto, and G. N_ McDermott. 1965.
Summary report on the effect of heavy metals on the biological
treatment processes. Journal Water Pollution Control Federation
37(l):86-96.
Berry, E. C. 1966. Requirements for microbial reduction of farm
animal wastes. Proceedings of the National Symposium on Animal
Waste Management Publication No. SP-0366, American Society of
Agricultural Engineers. St. Joseph, Michigan.
Bosch, H. M., A. B. Rosenfield, R, Huston, H. R. Shipman, and R. L.
Woodward. 1950. Methomoglobinemia and Minnesota well supplies.
American Water Works Association Journal 42:161-170.
Bridgham, Do 0. and J. T. Clayton. 1966. Trickling filters as a dairy
manure stabilization component. Proceedings of the National
Symposium on Animal Waste Management Publication No. SP-0366,
American Society of Agricultural Engineers. St. Joseph, Michigan.
Clark, C. E. 1965. flog waste disposal by lagooning. Journal of the
Sanitary Engineering Division, American Society of Civil Engineers
91(6):27-41.
Collins, V. G. 1963. The distribution and ecology of bacteria in
freshwater. Proceedings of the Society for Water Treatment and
Examination 12(l)s40-66.
Curtis, D. R. 1966. Design criteria for anaerobic lagoons for swine
manure disposal. Proceedings of the National Symposium on Animal
Waste Management Publication No, SP-0366, American Society of
Agricultural Engineers, St. Joseph, Michigan.
Decker, W. M. and J. H. Steele. 1966. Health aspects and vector control
associated with animal wastes.- Proceedings of the National Symposiu
on Animal Waste Management Publication No. SP-0366, American Society
of Agricultural Engineers. St. Joseph, Michigan.
97
um
-------
Environmental Pollution Panel. 1965. Restoring the quality of our
environment. Report of the President's Science Advisory Committee,
The White House, Washington, B. C,
Fair, G. M. and J. C. Geyer. 1954. Water Supply and Waste-water
Disposal. John Wiley and Sons, Inc. New York.
Fish, H. 1964. Pollution prevention in relation to agricultural
effluents. Chemistry and Industry 9:354-356.
Ford, D. L. 1968a. Application of the total carbon analyzer for
industrial wastewater evaluation. Proceedings on the 23rd
Industrial Waste Conference. Purdue University Engineering Exten-
sion Services No. 132, pp. 989-999.
Ford, D. L. 1968b. Total organic carbon as a wastewater parameter.
Public Works 99(4);89-92.
FWPCA. 1969. FWPCA methods for chemical analysis of water and wastes.
Analytical Quality Control Laboratory. Federal Water Pollution
Control Administration. U. S. Department of the Interior.
Cincinnati, Ohio.
Geldreich, E. E. 1966. Sanitary significance of fecal coliforms in the
environment. Publication WP-20-3, Control Research Series, Federal
Water Pollution Control Administration, U. S. Department of tke
Interior. Washington, D. C.
Geldreich, E. E., H. F. Clark, C. B. Huff, and L. C. Best. 1965. Fecal
coliform organism medium for the membrane filter techaique.
American Water Works Association Journal 57(2):20S-214.
Goleuke, C. G. and W. J. Oswald. 1966. Treatment and reclamation of
agricultural waste waters. Proceedings of the Symposium on
Agricultural Waste Waters, Report No. 10 Water Resources Center,
University of California, Davis.
Grub, W., R. C. Albin, D. M. Wells, and R. Z. Wheaton. 1969. The
effect of feed, design, and management on the control of pollution
from beef cattle feedlots. WRC-69-3 Water Resources Center, Texas
Technical College, Lubbock.
Hartman, R. C. 1965. Is poultry manure marketable? Poultry Tribune
(January):44.
Hemingway, R. G. 1961. Mineral composition of farmyard manure.
Emperial Journal of Experimental Agricultural 29:14-18.
Henderson, J. J. 1962. Agricultural land drainage and stream pollution.
Journal of the Sanitary Engineering Division, American Society of
Civil Engineers 88(6):61-73.
98
-------
Herron, G~ ML and A. B. Erhart, 1965, Value of manure on an irrigated
calcareous soil. Soil Science Society of America Proceedings
29:278-281.
Howells, D, Ho L969., Animal waste, problem mounts. The Enterprise.
Wilmington, North Carolina (September I).
Irgens, R, L. and D, L., Hay, 1966. Aerobic treatment of swine wastes „
Proceedings of the National Symposium on Animal Waste Management
Publication No. SP-0366, American Society of Agricultural Engineers.
St. Joseph, Michigan,
Jenkins, D.. 1967., Analysis of estuariue waters. Journal Water Pollution
Control Federation 39(2):159-180.
Jones, J. R. 196&. North Carolina swine situation. Prepared from
Estimated Swine Situation Reports by County Extension Workers,
Extension Anirj-al Husbandry, North Catol.ifia'State University at
Raleigh.
Kabler, P. W. and H. F. Clark. 1960, Colifonn group and fecal coli-
form organisms as indicators of pollution in drinking water.
American Water Works Association Journal 52s1577-1579.
Kamphake,- L. J. , S, A. Hannah, and J, M, Cohen, 1967. Automated
analysis for nitrate by hydrazine reduction. Water Research
1:205-216.
Kenner, B, A., H. F. Clark,and P. ¥. Kabler. 1961. Fecal streptococci
I. Cultivation and enumeration of streptococci in surface waters.
Applied Microbiology 9s16-20.
Levin, G. V. 1967. The new pollution. Civil Engineering. 5:68-71.
Loehr, R. C. 1967. Effluent quality from anaerobic lagoons treating
feedlot wastes. Journal Water Pollution Control Federation 39(3);
384-391,.
Loehr, R. C. and R, W. Agnew. 1967. Cattle waste-pollution and potential
treatment. Journal of the Sanitary Engineering Division, American
Society of Civil Engineers 9-3 (4) s 55-72.
Ludwig, H. F. , A. L. Gram, and J. L. Feeney. 1961. Pollution - urban
vs rural in Imperial. Valley waters - Part I and Part II. Water
and Sewage Works 108(7>:251-254 and 108(8)^308-314.
Lundgren, D« P, 1960. Phosphate analysis with the Technicon Auto-
Analyzer, Analytical Chemistry 32(6):824-828.
McHenry, S. 1961 What is poultry manure worth? Compost Science
2(3):13-15.
99
-------
Meitnan, J. R. and So H. Eumkle. 1966o Land treatment and water
quality control. Paper presented at the 21st Annual Meeting of
the Soil Conservation Society of America. ALburquerque, New
Mexico,,
Miner, J. R., L. R. Fina, J. W. Funk, R. I. Lipper and G. H. Larson.
1966a. Stormwater runoff from cattle feedlofcs. Proceedings of
the National Symposium on Animal Waste Management Publication
No. SP-0366, American Society of Agricultural Engineers. St.
Joseph Michigan.
Miner, J. R., R. I. Lipper, L. R. Fina, and J. W. Funk. 1966b. Cattle
feedlot runoff - its nature and variation. Journal Water Pollut-
ion Control Federation 38:1582-1591.
Morrison, S. M., D. W. Grant, Sister M. P. Nevins, and K. Elmund. 1969.
Role of excreted, antibiotic in modifying microbial decomposition oi
feedlot waste. Animal Waste Management. Cornell University
Conference on Agricultural Waste Management. Ithica, New York.
pp. 336-339.
Meuhling, A. J. 1969. Swine housing and waste management: a research
review. AEng - 873, Department of Agricultural Engineering,
University of Illinois. Urbana - Champaign, Illinois.
NCDA. 1968a. North Carolina agricultural statistics; 1968 issue.
No. 119, Cooperative Crop Reporting Service, Raleigh, North
Carolina.
NCDA. 1968b« North Carolina agricultural statistics; livestock, dairy
and poultry issue; historical data 1867-1968. No* 118. Cooperative
Crop Reporting Service, North Carolina Department of Agriculture,
Raleigh, North Carolina.
NCDA. 1970. North Carolina farm report. No. 551. Cooperative Crop
Reporting Service. Raleigh, North Carolina.
O'Brien, J. E. and J. Fiore. 1962. "Robote chemist" determines
nitrates in sewage and wastes. Waste Engineering 33(3):128-131.
Philps, A. 1965, Cageman's fringe benefit - dried manure? Poultry
Tribune (February):18.
Placak, 0. R. , C. C. Ruchhoft, and R. G. Snapp. 1949. Copper and
chromate ions in sewage dilutions, effect on BOD, Industrial
and Engineering Chemistry 41:2238-2241.
Rainwater, F. H, and L. 1L, Thatcher. 1960. Methods of collection
and analysis of water samples. Water Supply Paper 1454. United
States Geological Service, United States Government Printing
Office. Washington, D. C.
100
-------
Reed, C. H. 1966. Disposal of poultry manure by plow - furrow - cover
method. Proceedings of the National Symposium of Animal Waste
Management Publication No. SP-0366, American Society of Agricultural
Engineers. St. Joseph, Michigan.
Robbins, J. W. D. and G, J. Kriz. 1969. Relation of agriculture
to groundwater pollution: a review. Transactions of the ASAE
12(3):397-403.
Robbins, J. W. D. and G. J. Kriz. 1970. An automatic liquid sampler.
Prepublication article. Department of Biological and Agricultural
Engineering, North Carolina State University at Raleigh.
Robbins, J. W. D. , G. J, Kriz, and D. H. Howells, 1969a. Total
organic carbondeterminations on swine waste effluents. Paper
No. 69-928, American Society of Agricultural Engineers. St
Joseph, Michigan.
Robbins, J. W. D. , G. J. Kriz, and D. H. Howells. 1969b, Quality of
effluent from swine production areas. Paper No. 69-706, American
Society of Agricultural Engineers. St. Joseph, Michigan.
Sawyer, C. N. 1960. Chemistry for Sanitary Engineers, McGraw-Hill,
New York.
Schaffer, R. B. , C . E. Van Hall, G. N. McDermott, D. Barth, V. A.
Stenger, S. J. Sebesta, and S. H. Griggs. 1965. Application
of a carbon analyzer in waste treatment. Journal Water Pollution
Control Federation 37(11):1545-1566,
Secretary of Agriculture, 1969. Control of agriculture-related
pollution. Report to the President, The White House. Washington,
D. C.
Smith, S. 1965, Pollution-caused fish kills in Kansas. Kansas Water
News 8(2);8.
Sobel A. T. 1966. Physical properties of animal manures associated
with handling. Proceedings of the National Symposium on Animal
Waste Management Publication No.. SP-0366, American Society of
Agricultural Engineers. St, Joseph, Michigan.
Taiganides, E. ~E, 1963. Characteristics and treatment of wastes
from a confinement hog production unit. Unpublished Ph.D.
thesis, Department of Agricultural Engineering, Iowa State
University, University Microfilms. Ann Arbor, Michigan.
Taiganides,- E. P. and T. E. Hazen. 1966. Properties of farm animal
excreta. Transactions of the ASAE 9(3):374-376.
101
-------
Taylor, E. W. , N. P. Burman, and C. V. Oliver. 1955. Membrane
filtration technique applied to the routine bacteriological
examination of water* Journal of the Institute of Water
Engineers 9:248-254.
Technicon Corporation. 1965., Kjeldahl nitrogen. Bulletin-N-3, Auto-
Analyzer Methodology, Tarrytown, New York.
Technicon Corporation. 1969. Phosphate (ortho) in water and waste
water (range 0-400 ppm). Industrial Method 2-68W, Auto-Analyzer
Methodology. Tarrytown, New York.
Toth, S. J. 1965. Agricultural values of dried poultry manure and
bedding. Compost Science 5(3):29-30.
Tower, B. A. and H. T. Barr. 1965. Will indoor lagoons work in the
South? Poultry Tribune (February):35.
Van Hall, C. E. and V. A. Stenger. 1964. Use of infrared analyzer
for total carbon determination. Water and Sewage Works 111(6):266-
270.
Van Hall, C. E. and V. A, Stenger. 1967. An instrumental method for
rapid determination of carbonate and total carbon in solutions.
Analytical Chemistry 39:503-507.
Wadleigh, D. H. 1968. Wastes in relation to agriculture and forestry.
Miscellaneous Publication No. 1064, United States Department of
Agriculture, United States Government Printing Office, Washington,
D. C.
Webster, N. W. and J, T. Clayton. 1966. Operating characteristics of
two aerobic-anaerobic dairy manure treatment systems. Proceedings
of the National Symposium on Animal Waste Management Publication
No. SP-0366, American Society of Agricultural Engineers. St.
Joseph, Michigan.
Weibel, S. R., R. Bo Weidmer, J. M. Cohen, and A. G. Christiansen.
1966. Pesticides and other contaminants from rainfall and runoff
as observed in Ohio. Paper presented at the 1965 Annual Meeting
of the American Water Works Association. Portland, Oregon.
Wheatland, A. B. and B. J. Borne. 1964. Treatment of farm effluents.
Chemistry and Industry No- 9 :357-362.
Willrich, T. L. 1966. Primary treatment of swine waters by lagooning.
Proceedings of the National Symposium on Animal Waste Management
Publication No. SP-0366, American Society of Agricultural Engineers
St. Joseph, Michigan.
102
-------
Witzel, S. A., E. McCoy, L. B. Polkowski, 0. J. Attoe, and M. S. Nichols,
1966. Physical, chemical and bacteriological properties of farm
wastes (bovine animals). Proceedings of the National Symposium on
Animal Waste Management Publication No. SP-0366, American Society
of Agricultural Engineers. St. Joseph, Michigan.
103
-------
APPENDIX
104
-------
Appendix A
Automatic Sampler
An automatic sampler developed to collect the effluent samples for
the study is presented in Figures 30, 31, and 32 (Robbins and Kriz, 1970).
Advantages of the sampler include D.C. operation, reliability, simplicity,
and economical construction. Features to satisfy specific requirements
of sampling are easily incorporated into the basic design. Samples
collected with the automatic sampler showed no significant difference
from grab samples in content of nutrients (Kjeldahl, ammonia, nitrate
and nitrite nitrogen, and total and orthophosphate), "organics" (biochem-
ical oxygen demand, chemical oxygen demand, total organic carbon, total
solids and volatile solids), bacteria (total coliform, fecal coliform
and fecal streptococci), pH, or specific conductivity.
The sampler has the capability to collect and store water samples
from intermittently or continuously flowing streams. Up to twenty-four"
individual quart samples can be collected on a variable time basis.
Each narrow mouth mason jar is fitted with an #11 1/2 rubber stopper
that has a 1/16-inch air vent and a 1/4-inch sample inlet. The inlet
is fitted into a 1/4-inch diameter teflon or polyethylene tubing through
which the sample enters the jar. The jars are stored on an expanded
metal tray inside an insulated box that is made of 2-inch styrofoam
covered with sheet metal. The container keeps ice for several days.
When the water level in a stream raises the float (31-1) suffi-
ciently to close the microswitch (31-2), the sampling sequence begins.
The sampling time interval is preset by choosing the desired combination
of time of revolution of the chart drive (31-3) and number of notches on
the chart drive drum. When the chart drive closes the microswitch (31-4),
the pump (31-5) and solenoid (31-6, 32-6) are activated. The solenoid
(31-6, 32-6) orients the terminal end of the 5/8-inch pump tube to one of
the sample inlet tubes by rotating the pull wheel (32-7, 33-7) 1/12 of a
revolution. This moves the pump tubing holder (32-8, 33-8) one inch
horizontally. The time delay relay (31-9) breaks the electrical circuit
after sufficient time is allowed to collect the sample. This prevents
the battery from discharging needlessly and the solenoid from overheating.
Then, the plunger return spring (32-10) moves the solenoid plunger to
the next notch on the pull wheel (32-7, 33-7). As the chart drive (31-3)
continues to revolve, the microswitch (31-4) opens. This allows the
time delay relay (31-9) to close and the sequence for taking the next
sample begins when the microswitch (31-4) is again closed by the chart
drive. When the water level drops sufficiently to close the microswitch
(31-2), the sampling sequence stops.
The friction spring (33-11) regulates the pull necessary to rotate
the pull wheel (32-7, 33-7) and to prevent the wheel from rotating back-
wards when the solenoid plunger is pulled by the plunger return spring
(32-10). The chart drive (31-3) may also be used to record the water
level in the stream.
"''Two numbers in parenthesis refer to specific Figure-Component.
105
-------
12 VDC Solenoid
Cromar # SL7D2844 (6)
12 V Auto
Battery
12 VDC Submersible
Pump - Crainger
# 1P6A6 (5)
\ 60 Sec. Time Delay
\JRelay - Amprite
# 12C60 (9)
1 Microswitch
Normally closed (2)
Microswitch
BZ-2RW80-A2
Float
(1)
(4)
Weather Bureau Rain Gage or
Water Level Recorder - 8-Day
Chart Grive with Knotches
Added to Activate Switch (3)
Figure 3Q. Circuitry for automatic sampler
106
-------
Wire Rope
(fishing line)"
Guide Wheel-
Sample Inlet Tube,
Putnp Tubing
Holder (8)
Solenoid Plunger
Solenoid(6)~\
Pull Wheel (7)
Plunger Return Spring (10)
UI) O (PTJQ O O Q GD O O CTGIJIJOTXJOTKJ
1"
14"
-30"
Figure 31. Advancing mechanism for automatic sampler
-------
o
00
Pull Wheel (7;
Friction Spring (11;
Lid
Pump Tubing Holder (8)
12 Notch
Wheel
- Wire Rope Take-up
1—Wire Rooe Let-Out
Sample Inlet Tube
11 1/2
Rubber
Stopper
Jar
Sheet
Metal
Expanded Metal Tray
Styroroam
it
5/8" Pump Tube
To Pump
Figure 32. Cut-away view of automatic sarplcr
-------
Cost of materials excluding the chart drive is less than one
hundred dollars and the labor required for construction does not exceed
two man days. Useful modifications include use of an additional solenoid
for added pull (connected in parallel electrically), separate circuits
for the pump and solenoid, and different size sample containers.
109
-------
Appendix B
Illustrative pages from the Supplemental Appendix reporting the
following:
Raw data
Simple statistics by data groups
Regression coefficients
110
-------
Illustrative Page - raw data
Poultry Study Data
The key to the listing in Table 25 is the same as for
Table 23 except for changes as follows:
Column
20-27
53-56
61-63
Table 25. Listing
by data
m d h d. d.
t b
DATA GROUP XI
2 214 1 0
239 20
DATA GROUP X2
216 9 40 G
2161C 46 C
21611 52 0
21612 50 G
21613 49 0
21615 48 C
21616 47 C
21618 60 0
2162C 72 C
21622 90 0
21624 78 C
217 2 70 0
217 4 64 G
217 6 53 C
217 8 46 0
21710 42 G
21712 37 G
21716 25 G
21718 18 C
DATA GROUP X3
331 4 31 C
«• *rf 4* « •* • ^
331 8 1C 0
«• «V •• W * "" ^
4 116 16 0
4 118 12 G
^^ A 4fe ^£ •• ••* ^
A 2 1 6 C
•^ fc • ^"* **
4 2 9 15 0
Data
COD
Blank
Blank
Units
10" Wl
-
-
of effluent quality and hydrological data
groups ,
FC
IOC
174
100
80
100
1C50
510
200
170
29G
180
230
SO
6C
poultry study
COD BODC TOC
5
520 120 370
330 75 280
124001070010660
90801000010620
11250 82G011080
14800 3dOO 6780
14250 6500 6640
S250 4800 4180
17200 5200 56CO
6740 3060 2950
5400 1300 2080
5770 1400 2430
4930 1080 2060
3410 700 1100
3220 620 1000
2730 620 810
2790 740 860
2540 680 850
2760 860
4620 950
4530 1020
5200 190 390
2320 360 450
4950 250
2470 156
23>90 ITO
TN NHn NO, TP
3 3
1 5
1325
2700 6
2400 45
1930 45
1330155
1350 30
850126
1210 20
670 48
350123
260151
200160
180227
150238
180368
250370
220350
300313
240350
3202<35
12160
10250
12130
7235
9375
7485
T T
x n
5829
6648
6346
6346
6346
6346
6346
6346
6346
6346
6346
6346
6346
4832
4832
4832
4832
4832
4832
4832
4832
3834
3834
6237
6237
6848
6848
111
-------
Illustrative Page - raw data
Table 96. Conductivity, pH, COD, TS, VS and FS data by site and date
Site
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
mo
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
11
11
11
11
11
11
11
11
11
11
1
2
2
2
2
2
2
3
4
Time
day
13
13
14
14
14
14
13
19
19
19
19
20
20
20
20
21
21
21
21
21
22
22
23
23
30
5
6
18
21
20
22
23
22
18
25
22
27
27
18
24
24
25
25
29
Conduc-
. L. J ..j i — _rt
••— " civicy, pit
hr ohms
21
22
3 7.1
6
8 7.4
9
18
12 7.3
16
20
24
4
8
12
18
11
12
14
16
12
14
22
12
14
5
12
12
12
13
9 5.3 7.2
12
9
10
20 5.2 7.4
10 6.0 7.3
10
10
12
20 7.4
12 5.2 7.3
24 6.0
10 5.5
16
9 46.1 7.9
COD TS
mg/1
2570
2540
2570
2570
2330
2280
2370
2560
2300
2170
2110
2060
2190
2160
2090
2160
2140
2150
2220
2150
2170
2180
2150
2070
1980
1960
1230 1890
1940
1830
1210
950 1250
1050
980
1200
1100
VS FS,
col/ 100 ml
1460
1460
1450
1460
1180
1130
1316
1620
1310
1410
1470
1310
1530
1510
1400
1150
1150
1060
1250
1230
1070
1120
1130
1120
110,000
100,000
60 ,000
840
970
850
850
810
650
590
370
82,000
1,400,000
160,000
390
600
112
-------
Illustrative Page
Table 27 . Simple statistics for FC paired with other data by data groups
6
R
0
D
TC
BOD 5
TOG
* col/ 100 ml
ABC
A
Al
All
A12
A2
1262640
920434
684*
197
1254095
678675
521'*
105
1328764
663534
485
89
1648545
556794
472
55
811471
469714
192
34
838750
627247
829
16
378
136
-442*
198
384
88
096
105
393
67
-175-
86
371
65
-146
51
424
59
133
35
347
149
395
19
366
97
037
135
396
67
-023
54
436
22
-222
35
0
436
22
-222
34
323
62
178
19
TN N
M
H3-N NQ.,-]
/ a / r x
W TP-P04 i
103 /
OP~P04
mg/1
222
65
-430*
178
243
44
156
100
255
33
119
85
279
19
-452*
51
219
8
-337*
34
176
33
_L • .
527**
15
173
75
-489*
104
187
41
242
59
203
28
284**
45
223
18
024
26
175
10
178
19
138
386
^L Jt
535
14
126
251
-186*
189
140
17
-036
105
144
14
082
89
153
9
-454*
55
131
11
101
34
119
14
-476
16
109
239
035
103
118
21
-006
70
124
18
087
57
128
18
-009
43
113
10
261
14
92
9
-321
13
No,
**
Tn
"P
59
13
-247*
206
65
12
-025
111
65
11
330*
90
71
7
167
55
56
11
218
35
63
16 *
-700
21
38
13
-181*
206
43
13
,171
111
43
13
480*
90
50
10
418*
55
32
10
324
35
44
15.
-524
21
H
206
102
-202*
206
300
000
000
111
300
000
000
90
300
000
000
55
300
000
000
35
300
000
000
21
qt
.015
.026
-050
206
.027
.030
215**
111
.027
.033
283*
90
.019
.030
585*
55
.040
.032
146
35
028
014
002
21
qb
cfs
.003
.006
-034
206
.005
.008
130
111
.002
.002
303*
90
.003
.002
199
55
.001
000
000
35
.020
•°10**
507**
21
qo
.012
.025
-042
206
.022
.031
177
111
.025
.033
266*
90
.016
.029
584*
55
.039
.032
146
35
009
.012
-391
21
-------
Illustrative Page
Table 81. Regression equation coefficients for FC a f(Tn. H, qo),
col/100 ml, by data groups
Group
ABC
A
All
A12
A2
A23
A 24
A25
B
Bl
B2
B22
B23
C
Cl
C2
C22
D
Dl
D2
D3
D4
1
814753.2
236361.5
404942.4
584939.3
-182311.0
-621116.5
3141196.3
-55930.1
1798357.6
2034901.0
-1982044.9
-5705196.7
-2462572.5
2649107.8
2286429.6
-768689.3
4144672.2
1585895.2
-364441.4
11190679.0
982076.4
111349657.3
•r Op
An* r
-3989.7
7351.6
253.5
-871.9
28834,1
41709.5
-64120,9
30784.4
-21057.8
-27922.0
106866.1
227615.4
151768.5
-8821. 4
8027.0
47421.6
-110631.2
-81329.6
-6110.3
-193961.5
1220.7
-1571491.7
H qos cfs
980020.8
-735129.4
3746235.4
-2754365.2
-28077396.9
17659091.7
18660264.9
-35972368,7
-71725211.0
251603898.9
-550226847.6
-1060629750,3
-914833474.8
-105681555.7
24275708.9
5703V555.8
24707.0 -24737.3
6395.6 -326348/9
1936064.2
-205677.6
174500.9
R2
.055
.068
.341
,061
.404
.486
.535
.810
.192
.333
.376
-798
,658
.005
.001
.119
.141
.052
.575
.201
.004
.133
No.
205
111
55
35
21
8
5
8
49
34
13
8
5
46
3 '3
8
6
88
8
25
26
29
114
-------
Accession Number
W
Subject Field & Group
056
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Organization
Department of Biological and Agricultural Engineering
School of Agriculture and Life Sciences
North Carolina State University at Raleigh
Title
Role of Animal Wastes in Agricultural Land Runoff
10
Authors)
Robbins, Jackie W. D.
Howells, David H.
Kriz, George J.
16
Project Designation
EPA/eiM Grant No. 13020 DGX 08/71
21
Note
22
Citation
*Animal Wastes, *Farm Wastes, *Runoff, *Agricultural Wastes, *Water Pollution,
*Farm Lagoons
25
Identifiers (Starred First)
*Animal Wastes, *Agricultural Land Runoff
27
Abstract
Twelve typical agricultural areas representing three types of animal waste
management techniques—lagooning, direct discharge into streams and land
spreading including pasture and drylot units—were studied to determine the amounts
of and factors governing stream pollution from swine, dairy, beef, and poultry
production operations.
Moee than 1500 stream and lagoon effluent samples were collected with an
automatic sampler developed for the study. The samples were analyzed for
bacteria, nutrients, and degradable organics. Hydrological and waste
management data were also collected.
Study results point to the superiority of land spreading for the disposal of
amimal wastes. Good soil and water conservation practices should be used to
minimize the movement of wastes into streams. Higher rates of runoff result
in heavier pollution. The location of disposal areas away from streams is
important in controlling the amount of entering wastes. Even when land
disposal areas are poorly located, the amount of pollution entering streams is
usually low; and watershed factors, such as surface culture and ease of erosion,
are of primary importance in governing the magnitude of pollution which reaches
the streams. Direct dumping of animal wastes, treated or untreated, into streams
is completely unacceptable and should be prohibited.
Abstract°6av±d H. Howells
Institution
North Carolina State University at Raleigh
WR:102 (REV. JULY 1969)
WRSIC
SEND WITH COPY OF DOCUMENT. TO: WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. D. C. 20240
* GPO! 1970—389-930
------- |