Constructed wetlands to treat wastewater from dairy and ... wetlands to tre… · Constructed wetlands to treat wastewater from dairy ... are removed from the water by plant uptake

ELSEVIER Agriculture, Ecosystems and Environment 58 (1996) 97-114

Agriculture Ecosystems & Enwronment

Review Article

Constructed wetlands to treat wastewater from dairy and swine operations: a review

Julie K. Cronk 1 Department of Agricultural Engineering, Uniuersity of Maryland, College Park, MD 20742, USA

Accepted 16 December 1995

Abstract

Animal wastewater can be a major contributor to the cultural eutrophication of surface waters. Constructed wetlands are under study as a best management practice to treat animal wastewater from dairy and swine operations. Preliminary results are promising when wetlands are a component of a farm-wide waste management plan, but they are ineffective without pretreatment of the wastewater. The feasibility of constructed wetlands varies with waste characteristics and climate. While the cost of wetland construction is low, the site must be maintained in order for the initial investment in the wetland to be worthwhile. In addition, several design iterations may be necessary before effective treatment is obtained. The design of animal wastewater treatment wetlands is still being researched and a number of the present projects will help provide recommendations for the use of constructed wetlands at animal operations.

Keywords: Animal wastewater treatment; Best management practice; Constructed wetlands; Dairy parlor effluent; Swine wastewater management; Waste management system

1. Introduction

Runoff from animal farms where excessive nutrients (i.e. quantities far in excess of plant nutrient needs) are generated has been linked to downstream eutrophication of surface waters (Jaworski et al., 1992). In the UK, a stream which drains a 400 ha watershed with seven dairy operations showed sig- nificant increases in ammonia and biochemical oxygen demand after rain (Schofield et al., 1990). In addition, biological monitoring of benthic macroin- vertebrates revealed significantly decreased populations of pollution intolerant species downstream of the farms.

J Present address: 3400 2nd St. Rt. #3, Wayland, MI 49348, USA. Tel.: 301-405-1198; fax: 301-314-9023.

The Monocacy River in northwestern Maryland has been identified as a major source of nonpoint source pollution to the Chesapeake Bay. The Mono- cacy watershed encompasses the greatest percentage of dairy cows of all Maryland watersheds (32%) and it produces the greatest amount of animal manure (11% of Maryland's total; Brodie, 1988). A study of the Upper Potomac River Basin (into which the Monocacy flows) estimated that 43% of the nitrogen input and 60% of the phosphorus input to the basin came from animal waste and that the Upper Potomac received twice the amount of nitrogen and phosphorus in the form of fertilizer and animal waste than required by the agricultural crops grown in the area (Jaworski et al., 1992).

In a long term study in North Carolina, Hunt et al. (1995) sought to determine the effect of animal

0167-8809/96/$15.00 Copyright © 1996 Elsevier Science B.V. All fights reserved. PII S 0 1 6 7 - 8 8 0 9 ( 9 6 ) 0 1 0 2 4 - 9

98 J.K, Cronk / Agriculture, Ecosystem and Environment 58 (1996) 97-114

waste on the water quality of a 2044 ha watershed. Land use in the watershed was agricultural and commercial fertilizer was used along with swine and poultry waste. The area's streams were closely con- nected to its shallow ground water and both surface and ground water contamination were a problem. A mass balance of nitrogen and phosphorus revealed that 8-26% of the nitrogen and 2-3% of the phosphorus were lost to the stream. The watershed was divided into three subwatersheds for the study and both stream and ground water nitrate levels were highest in the subwatershed with the highest level of animal waste production.

Animal waste disposal is of increasing concern because of larger and more concentrated livestock operations with less land per animal for manure application (Kashmanian, 1994). Further, as urban areas encroach on rural land, greater demands are placed on water resources. Since animal waste management is tightly constrained by economics, it is necessary to develop inexpensive and sustainable management practices which require low energy inputs. A number of best management practices target animal waste problems. These include aerobic and anaerobic lagoons, storage pits, land spreading, protective fencing, and composting. Recently, constructed wetlands have been proposed as an alternative wastewater treatment component and their effectiveness is under study at a number of sites.

2. Water quality improvement mechanisms in wastewater treatment wetlands

Nutrient and solids removal in wetlands is facili- tated by shallow water (which maximizes the sediment to water interface), high primary productivity, the presence of aerobic sediments and anaerobic sediments, and the accumulation of litter (Mitsch and Gosselink, 1993). In addition, slow water flow causes suspended solids to settle from the water column in wetlands. Biochemical oxygen demand (BOD) is reduced by the settling of organic matter and through the decomposition of BOD causing substances. Be- sides solids and BOD, the most important constituents in animal wastewater are nitrogen and phosphorus and these can both be reduced in constructed wetlands if conditions are appropriate.

Nitrogen enters an animal wastewater treatment wetland in either an organic or inorganic form. The inorganic forms are nitrate (NO3), nitrite (NO~-), ammonia (NH 3), and ammonium (NH~). Ammonia may be lost from the system through volatilization, taken up by plants or microbes, or oxidized to nitrate in the nitrification process. Similarly, ammonium may be taken up in the biota or nitrified. In addition, because of its positive charge, it can be sorbed onto negatively charged soil particles. Nitrate and nitrite are removed from the water by plant uptake or denitrification (Gambrell and Patrick, 1978). Once nitrogen has been denitrified, it is released to the atmosphere as nitrous oxide (N20) or dinitrogen gas (N2). Denitrification brings about a removal of nitrogen from the aqueous system and it is the most important removal pathway for nitrogen in most wetlands (Faulkner and Richardson, 1989). As the organic nitrogen is mineralized, it enters the inorganic nitrogen cycle. High nitrogen removal rates have been noted in many wetland studies (Johnston, 1991). Because the transformations of nitrogen in- volve microbial processes, nitrogen removal is enhanced during the growing season when high temperatures stimulate microbial population growth (Gambrell and Patrick, 1978). In addition, plant uptake of nitrogen only occurs during the growing season.

With the deposition of solids comes a decrease in contaminants which are associated with solids, such as phosphorus (Johnston et al., 1984). Bioavailable forms of phosphorus may be taken up by plants or sorbed onto soil particles. Adsorption of phosphorus is the main mechanism for phosphorus removal in wetlands (Richardson, 1985). Because phosphorus adsorption is enhanced under aerobic conditions, the removal of phosphorus in the anaerobic conditions of wetlands may be less than on dry soil (Richardson, 1985). Iron phosphates form under oxidized conditions and phosphorus may be released when the iron is reduced in hydric soils, therefore, caution should be taken in anticipating removal rates in wetlands when the wastewater contains large amounts of phosphorus. As phosphorus inputs to a constructed wetland continue over a period of several years, sorption sites in the sediments may become increas- ingly unavailable and eventually lead to phosphorus release (Kadlec, 1985). The removal of phosphorus

J.K. Cronk / Agriculture, Ecosystem and Environment 58 (1996) 97-114 99

through plant uptake and the decomposition of organic phosphorus increase with temperature. Grow- ing season increases in phosphorus retention have been observed at domestic wastewater treatment wetlands (Gearhart et al., 1989).

A number of strategies to enhance phosphorus sorption have been adopted at constructed wetland sites. These include distributing the inflow along the length of a pipe to increase the surface area of initial contact between the wastewater and the sediments (Hammer, 1992), and spraying the waste into the air over the wetland in order to oxygenate the waste (Swartz, personal communication). Adding aluminum to the substrate can also enhance phosphorus removal (James et al., 1992) since phosphorus adsorption is positively correlated to aluminum content in the substrate (Richardson, 1985). Periodic drain- ing can allow oxidation and recharge sorption sites for greater phosphorus removal than under permanently reduced conditions (Faulkner and Richardson, 1989).

Wastewater, both human and animal, may be contaminated with pathogens. Most related diseases in North America are caused by bacteria and viruses rather than worms and protozoa (Gersberg et al., 1989), so the treatment wetlands literature deals primarily with these two groups of organisms. Total or fecal coliforms are generally the only measured pathogen indicators in wastewater treatment wetlands. Coliforms are reduced within wetlands for a number of reasons: exposure to sunlight, predation, competition for resources, and toxins. In addition, they may be buried beneath sediment or adsorbed (Gersberg et al., 1989). In two cases in which constructed wetlands were used as tertiary treatment systems for domestic wastewater, bacterial and viral indicators were 90-99% removed (Gersberg et al., 1989). If the wastewater has not been pretreated, additional disinfection through chlorination or exposure to ultraviolet radiation may be necessary. Pathogen data from animal wastewater treatment wetlands have not been reported.

Fig. 1. Handling options for the liquid portion of the waste from a dairy operation including the possible placement of constructed wetlands and recycled wastewater flow. Arrows indicate the flow of wastewater (adapted from NRCS, I992).

100 J.K. Cronk / Agriculture, Ecosystem and Environment 58 (1996) 97-114

3. Wastewater handling options

3.1. Dairy farms

A typical dairy operation has a flush after each milking. Milk is the major organic contaminant from the dairy parlor. It has a high 5-day BOD (about 100000 mg BOD 5 l - l ) because of its high fat and sugar content (Holmes et al., 1992). While most of the milk is not wasted, some portion of the milk enters the waste stream through rinsing transfer pipes, receivers, and storage tanks. Manure and feed are other organic contaminants from the milkhouse.

Other contaminants from dairy parlors may include detergent, which is usually alkaline and may contain phosphates, an acid wash (to neutralize the alkaline detergent), and water softener salts contain- ing sodium and chloride. The amount of waste from

the milkhouse depends on farm management procedures. From 20 to 40 1 of fresh water day- l per cow are used where flushing of wastes is not practiced and up to 600 1 fresh water day-1 per cow are used if manure flushing and cow washing are a part of the procedure (NRCS, 1992).

Wastewater may receive primary solids separation in a settling basin and then it may flow into a lagoon and /or be land spread. Constructed wetlands could replace or be downstream from a lagoon and they could replace or precede land spreading (Fig. 1). Treatment of dairy parlor waste sometimes includes a septic tank/drain field system, however septic tanks are not recommended because of the propen- sity for the formation of a bacterial slime which can clog soil pores in the drain system (Midwest Plan Service, 1987). If a septic tank is in place, a constructed wetland could receive and treat the liquid portion of the waste.

Fig. 2. Handling options for the liquid portion of the waste from a dairy operation including the possible placement of constructed wetlands and recycled wastewater flow. Arrows indicate the flow of wastewater (adapted from NRCS, 1992).

J.K. Cronk ~Agriculture, Ecosystem and Environment 58 (1996) 97-114 101

3.2. Swine farms 4.1. Surface flow wetlands

Swine waste may be stored temporarily 'under- house' beneath slatted floors. The waste may discharge into lagoons (Natural Resource Conservation Service, United States Department of Agriculture (NRCS, 1992)). Currently many swine waste management systems include temporary storage in anaerobic lagoons followed by land spreading. Runoff from the land is potentially detrimental to downstream waters and the odor of surface spread waste is objectionable if not incorporated immediately into the soil. Therefore, new and sustainable waste management strategies are needed (McCaskey et al., 1994). If the solids are separated from the liquids and treated separately, constructed wetlands could receive and treat the liquid portion of the waste from a swine operation (Fig. 2). Primary settling is necessary and treatment in a lagoon is desirable upstream of the wetland.

4. Wetland design methods

Most existing designs for animal wastewater treatment wetlands are based on estimated or measured BOD loadings. The equations used to calculate the desirable size and hydraulic residence time take into consideration an average decay rate of BOD, average temperature, and the hydrology of the site. The assumptions within the equations are derived from results from municipal wastewater treatment plants (Reed et al., 1988). An alternative is to base the design on nitrogen ioadings since many constructed wetlands reduce BOD, TSS, and bacteria concentrations, but nitrogen removal is variable (Hammer and Knight, 1994). Designs which feature a series of cells, each of which enhances one process within the nitrogen cycle, may be desirable to maximize nitrogen removal.

Before construction of an animal wastewater treatment wetland, selection of a suitable site includes consideration of the goals of the wetland, soil type, topography, climate, cost, and potential future changes in farm management (Mitsch and Cronk, 1992; McCaskey et al., 1994). For example, if the herd size is going to increase, the wetlands need to be large enough to handle the increased waste load.

Surface flow wetlands are the most common type used in the treatment of animal wastewater in the US. They are less expensive to construct than subsurface wetlands because fewer construction steps and materials are involved. The substrate in a surface flow wetland is usually compacted clay or hydric soils, if available. Influent waste flows over the substrate.

4.1.1. Natural Resources Conservation Service design guidelines

The design guidelines provided by the NRCS are based on municipal wastewater treatment wetlands and the experiences of those working with similar systems in the Tennessee Valley Authority (Reed et al., 1988). The NRCS cautions that the guidelines are technical requirements rather than a Conservation Practice Standard. The technology for wetland construction is still under study so the requirements allow more flexibility than Conservation Practice Standards for other waste management structures, such as lagoons (Krider and Boyd, 1992). An important factor in the design of NRCS wetlands for animal wastewater is that no discharge is permitted under normal operating conditions. Discharge into a waterway would require a permit based on effluent concentrations, however wetlands technology must be refined further before effluent concentrations that will meet permit requirements can be guaranteed. All wetlands built under these guidelines must contain a 25-year storm event. Since precipitation exceeds evaporation in areas where treatment wetlands are feasible, it may be desirable to build a detention pond downstream from the wetlands so that if an outflow does occur, the no-discharge requirement can still be met (McCaskey et al., 1994).

The NRCS recommends that wetlands for animal wastewater treatment be constructed as part of an overall waste management system and that the wetland influent be pretreated. Settleable and floating solids should be removed before they enter a treatment wetland. The design requirements for these wetlands are: 1. allowable BOD 5 loading rate to a constructed

wetland is 73 kg ha - l day-~ (65 lb acre -1 day-~);


2. the residence time must be at least 12 days. The BOD loading rate is conservative compared

with the recommendations of Reed et al. (1988), of a BOD loading of 112 kg ha- t day- l (100 lb acre- day- l ) . The residence time depends upon the average temperature and the length of time it takes to degrade the BOD.

The objectives of these systems are to improve water quality so that in the event of an outflow from the wetlands: 1. BOD 5 < 30 mg 1-~; 2. TSS < 30 mg 1-1; 3. N H s + N H ~ - N < 1 0 m g l - l .

The guidelines provide no standards for nitrate- nitrogen or for any other contaminants. The NRCS guidelines provide two methods for determining the size of a constructed wetland. Both methods are based on BOD 5 loadings. The presumptive method, assumes that livestock produce a certain amount of BOD 5 per day. This method is useful if the actual effluent can not be evaluated because the waste treatment system or one of the farm structures such as the dairy parlor has not been installed. The second NRCS method, known as the field test method, is based on measured BOD 5 concentrations. The NRCS recommends that several samples be collected during different seasons in order to obtain a good estimate of average BOD 5. Under both the presumptive and field test methods, inputs to the design equations are flow rate, porosity, water temperature, and either assumed or actual BOD 5 concentration.

Both design procedures are based on plug flow kinetics (Reed et al., 1988). The general model developed for the design of constructed wetlands is given below.

Ce/C o = a exp[( - 0.7K,( A v ) 1 . 7 5 L W d n ) ] / a ]

(1)

where C e is effluent BOD concentration (mg I - l ) , C O is influent BOD concentration (mg 1-1 ), A is the fraction of BOD which is not removed as settleable solids near the headworks of the system (as decimal fraction), K t is the reaction rate constant (day- l ; Kt=K2o (1.1) (T-2°), K20=0.0057 day- l ) , A~ is specific surface area for microbial activity (m 2 m-3) L is length of the system (m) W is width of the system (m), d is design water depth in the system

(m), n is porosity of the system (n = Vv/V where V v is volume of voids and V is total volume; expressed as a decimal fraction), and Q is average hydraulic loading on the system (m 3 day- l; Q = (Qinf luen t +

a e f f l u e n t ) / / 2 . 0 ) • The porosity of the wetland flow path (n) is often

taken as 0.75 and is based on measures of water flow through wetland plants. It is the ratio of the volume of constructed wetland occupied by water to the volume occupied by plants and water. The guidelines provide porosity values for different macrophyte species: 1. cattail (Typha spp.) = 0.95 2. bulrush (Scirpus spp.) = 0.86 3. reed (Phragmites communis) --- 0.98 4. rush (Juncus spp.) = 0.95.

An additional factor is introduced to account for the BOD which is removed by sedimentation. The value of A is often assumed to be 0.52, using empirical data from municipal waste treatment systems. This is considered the lower limit for A and it is not to exceed 0.90.

The hydraulic residence time t (in days) in terms of equation 1 is given by:

t = [(lnC o - lnCe) - 0 .6539]/65K, (2)

NRCS recommends that the overall length-to- width ratio be 3:1 to 4:1, with each cell having a 10:1 ratio. The 10:1 ratio may be altered depending on space available. To accomplish this, cells are generally arranged side by side across the width. The overall area needed for treatment is divided into cells to provide uniform flow down the length of the system as recommended by Reed et al. (1988). Ham- mer (1992) suggests a length-to-width ratio of 5:1 to reduce the tendency for excessive loading at the inlet of long narrow cells. Recommended depth varies from 0.15 to 0.20 m. Most of the animal wastewater treatment wetlands currently under study in the US were designed according to guidelines of NRCS (1991). At three dairy wastewater treatment wetlands built according to NRCS guidelines percent retention of BOD 5 was from 61 to 78% (Cronk et al., 1994; Reaves et al., 1994b; Skarda et al., 1994) and retention of other parameters at these sites was above 50% (Table 1).

Tab

le 1

M

ass

of c

onta

min

ants

ret

aine

d pe

r un

it a

rea

in w

etla

nds

for

anim

al w

aste

wat

er t

reat

men

t (g

m -

2 d

ay-

~).

Per

cent

red

ucti

on i

n m

ass

betw

een

infl

ow a

nd o

utfl

ow i

s gi

ven

in

pare

nthe

ses.

Val

ues

wer

e es

tim

ated

fro

m a

utho

rs'

repo

rts

of f

low

dat

a, w

etla

nd s

ize,

and

con

cent

rati

on

Loc

atio

n T

ime

peri

od

BO

D s

TS

S

TK

N

NH

3

TP

S

ourc

e (g

m-

2 d

ay-

t)

(g m

- 2

day

- t)

(g

m-

2 d

ay-

i)

(g m

- 2

day

- i )

(g

m -

2 d

ay-

t)

Dai

ry

Dro

into

n, U

K

Son

oma

Col

, C

A c

Fr

eder

ick

Co.

, M

D

LaG

rang

e C

o.,

IN

Ore

gon

Sta

te U

niv.

, OR

d

Swin

e A

ubur

n U

niv.

, A

Le

Low

loa

d (0

.23

g B

OD

s m

-2

day

i)

Med

ium

loa

d (0

.46

g B

OD

5 m

- 2

day

- i )

H

igh

load

(1.

1 g

BO

D 5

m -

2 d

ay-

J)

MS

Sta

te U

niv.

, M

S

Dup

lin

Co.

, N

C

Nov

.-A

pr.

a 3.

7 (1

7)

5.1

(57)

Ju

ne-S

ept.

b

24.3

(49

) 14

.3 (

70)

July

-Oct

. 7.

4 (9

7)

95.4

(99

) 4.

5 (9

6)

1.4

(86)

2.

1 (9

3)

Aug

.-A

pril

16

.5 (

70)

8.2

(90)

0.

6 (7

8)

0.3

(78)

0.

1 (5

4)

Jan.

-Dec

. 0.

9 (7

8)

1.9

(89)

0.

3 (6

6)

0.1

(66)

0.

03 (

64)

Oct

.-M

arch

8.

7 (6

1)

8.4

(73)

2.

1 (5

7)

1.4

(54)

0.

4 (6

6)

July

-Sep

t.

0.1

(95)

0.

3 (9

9)

0.2

(90)

0.

6 (9

6)

0.4

(72)

1.

3 (9

4)

Apr

il 1

992

0.3

(54)

0.

8 (6

9)

-Jul

y 19

93

June

-Dec

.

0.2

(99)

0.

2 (9

4)

0.2

(94)

0.

5 (9

7)

0.5

(98)

0.

3 (8

4)

1.0

(82)

1.

0 (8

9)

0.6

(79)

1.

2 (7

1)

0.2

(44)

0.3

(99)

Bid

dles

tone

et

al.,

1991

Cit

y of

San

ta R

osa,

199

4 C

ronk

et

al.,

1994

R

eave

s et

al.,

199

4a

Ska

rda

et a

l., 1

994

McC

aske

y et

al.

, 19

94

Cat

hcar

t et

al.,

199

4

Hun

t et

al.

, 19

94

t.,,i

a D

airy

par

lor

was

hing

s on

ly.

b S

ame

site

as

abov

e: d

airy

par

lor

was

hing

s + b

arny

ard

runo

ff.

c T

reat

s 21

% d

airy

was

tew

ater

and

79%

tre

ated

dom

esti

c w

aste

wat

er.

d A

ssum

es in

flue

nt f

low

rat

e eq

uals

eff

luen

t fl

ow r

ate.

e

Thr

ee l

oadi

ng r

ates

int

o th

ree

sets

of

two

cell

s in

ser

ies.


4.1.2. Marsh-pond-meadow A surface flow wetland that accommodates varied

depths may best promote nitrification and denitrification (Hammer, 1992). Components of the marsh- pond-meadow (or marsh-pond-marsh) surface flow wetland are as follows. 1. Marsh - -a shallow basin with dense emergent

vegetation such as cattail (Typha latifolia). The marsh's function is ammonification and the removal of BOD 5, TSS and pathogens.

2. Pond--constructed pond of 0.5-1 m depth which is similar to an aerobic lagoon. Vegetation includes surface species such as duckweed (Lemna minor) and submerged species such as pondweed (Potamogeton spp.). The pond's purpose is to further reduce BOD 5 and facilitate nitrification and denitrification.

3. Meadow (or shallow marsh)--overland flow system with reed canary grass (Phalaris arundi- nacea) or other inundation-tolerant grasses. The meadow receives effluent from the pond in shallow sheet flow having a depth from 1 to 5 cm. The meadow's purpose is to remove TSS and to further promote nitrification and denitrification. Each cell should have a length to width ratio of

3-5:1. Hammer (1993) reports on a marsh-pond-marsh

system constructed at the Pontotoc Experiment Sta- tion in Mississippi to treat swine wastewater. Each of the two cells has a 9-m-long shallow (12 cm) emergent marsh at the influent end. The cell then deepens to 35 cm for a 15-m-long pond followed by a final 9-m-long area 12 cm deep. The two cells are followed by two 46-m-long vegetated strips. During 16 months of sampling, ammonia removal was consis- tent and 71% of ammonia (by mass) was removed within the wetlands (Cathcart et al., 1994). Ammonia reduction may have been favored by the marsh- pond-meadow design which places an emphasis on nitrogen removal. When the post-wetland vegetated strips are included in a mass balance of measured parameters, the entire system removed over 95% of the incoming mass of BOD 5, TSS, NH3-N and TP (Cathcart et al., 1994).

A variation on this design type was constructed at Auburn University's Sand Mountain Agricultural Experiment Station in Alabama (Hammer, 1992). There, swine waste flows from a lagoon into a

mixing pond, then into five pairs of emergent marshes and finally into a wet meadow. Removal of TKN (82-99%) and N H 3 (89-99%) were relatively high, perhaps the result of the system's design focus on nitrogen removal (McCaskey et al., 1994; Table 1).

4.2. Subsurface flow wetlands

Subsurface flow wetlands are used extensively in Europe to treat human and animal wastewater. The substrate is usually gravel or some other coarse material. Subsurface flow systems are susceptible to clogging, therefore they are not recommended for waste with a high concentration of total solids (Hammer, 1994). They may be best suited to treat septic tank effluent and dilute chemical wastes (Mc- Caskey et al., 1994). The Tennessee Valley Author- ity is conducting research on subsurface flow constructed wetlands, investigating mechanisms of nitrogen, carbon, and phosphorus uptake in an effort to provide design and operating guidelines (Sikora, 1994).

In Europe, engineered reed beds are designed according to plug flow reactor kinetics. The beds are usually sealed with clay or a synthetic membrane and planted with reed (Phragmites australis). Sizing of these wastewater treatment beds is based on work by Kickuth in Germany (Biddlestone et al., 1991). The size is based on flow rate, inflow BOD concentration, desired outflow BOD, and a constant. A bed depth of 0.6 m is recommended at the inlet with a slope of 2-6%. Water depth is only 25 mm, so most of the flow is subsurface. It is desirable to construct reed beds on a slope to facilitate water flow. The area of the bed is calculated using the following equation:

a h = KlOd(lnC o -- lnCt) (3)

where A h is surface area of bed, KI is a constant, Qd is the daily average wastewater flow rate (m 3 day- I ), C O is average BOD of inlet wastewater (mg 1-~), and C t is the required average BOD of outlet water (mg 1- ~ ).

The constant K L depends on the biodegradability of the waste. For sewage, the value is 5.2; it should be determined experimentally for other waste. For both the Kickuth equation and the NRCS guidelines, the values for the constants were calculated for


domestic sewage. Applying them to design for wastewater from animal operations is based on the assumption that the value is still valid at BOD concentrations that may be an order of magnitude greater than for sewage.

Hydraulic residence time is calculated in the same manner as in the NRCS method using the following equation:

C, --- Coe-*' (4)

where k = 0.032 for sewage at 8°C, but must be determined for other wastes, and t is time (days).

5. The effectiveness of animal wastewater treat-

ment wetlands

Several natural and constructed wetlands have been tested for their ability to retain or transform nutrient inputs from municipal wastewater (Odum et al., 1977; Dierberg and Brezonik, 1983; Kadlec, 1987; Knight et al., 1987; Brodrick et al., 1988; Hammer, 1989). These have proven effective by decreasing nitrogen, phosphorus, suspended solids

and BOD. Owing to the successes in domestic wastewater treatment, constructed wetlands to treat animal wastewater have been included as a component of the waste management strategy on a number of private and research farms (Table 2). Constructed wetlands have had appeal as a farm waste management practice because they are low-cost, low-technology, and ideally, require little human-controlled energy input after construction (Hammer, 1992). However, BOD concentrations in livestock wastewater are usually at least an order of magnitude higher than in domestic wastewater. Therefore, although wetlands for secondary and tertiary sewage treatment have proven successful, designs adapted to the high BOD and nutrient concentrations of animal wastewater are necessary.

Most US animal wastewater treatment wetlands are in the southeastern states where precipitation exceeds evapotranspiration and moist conditions can be maintained. Warm temperatures and a long growing season in the southeast also favor the wetland functions that remove animal waste contaminants. Research in northern areas (Holmes et al., 1992) has

Table 2 Location and sizes of a subsample of the constructed wetlands used to treat dairy and swine wastewater at private farms and experimental stations. The sites are listed in chronological order of start date

Location No. of Total area Design type Start date Source cells (m 2)

Dairy Drointon, UK 1 250 Kickuth eqns. 1988 Newton Co., MS 6 205 NRCS 1989 Sonoma Co., CA 1 1046 WPCF 1990 1990 Dooly Co., GA 6 8700 NRCS 1992 Kosciusko Co., IN 2 2300 Modified NRCS 1992 LaGrange Co., IN 3 1116 Modified NRCS 1992 Calumet Co., WI 12 158 NRCS 1992 Frederick Co., MD 2 1092 NRCS 1993 Kent Co., MD 2 1214 NRCS: serpentine 1993 Oregon State Univ., OR 6 800 Modified NRCS 1993

Swine Auburn Univ., AL 6 2430 Marsh-pond-meadow 1988 MS State Univ., Pontotoc Co., MS 2 792 Marsh-pond-marsh 1991 Duplin Co., NC 6 720 NRCS 1992 Purdue Univ., West Lafayette, IN 16 580 Experimental plots 1993 Worcester Co., MD 6 13760 NRCS: serpentine 1993 Lincoln Univ., Cole Co., MO 12 567 NRCS 1995

Biddlestone et al., 1991 NRCS, 1989 Lanier et al., 1991 GSWCC, 1992 R.P. Reaves, personal communication Reaves et al., 1994b Holmes et al., 1992 Cronk et al., 1994 Baldwin and Davenport, 1994 Skarda et al., 1994

McCaskey et al., 1994 Hammer, 1993; Cathcart et al., 1994 Szogi et al., 1994 Reaves et al., 1994a Baldwin and Davenport, 1994 a Rastorfer and Schneider, 1994

a A serpentine design with one cell and four baffles was used to create six cells in series.


to take the effect of longer winters into consideration. There is no clear difference in the treatment of wastewater in warm and cold climates (Table 1) and it would be difficult to determine which climate is best suited to this treatment strategy since so many other variables exist. In addition to those wetlands listed in Table 2, a number of other dairy wetlands have been constructed, including several in Georgia, Alabama, the Virgin Islands, Puerto Rico (Surrency, personal communication), Maryland, (Davenport, personal communication), Maine (Holmes et al., 1995), and Louisiana (Chen et al., 1995). Other swine wastewater treatment wetlands exist as well and these include a number of sites in Georgia and Alabama (Surrency, personal communication). When constructed wetlands are located on private property, information concerning design, performance, and cost is often not easily available.

Some constructed wetlands have shown a high percent reduction in contaminant levels (Table 1); however, it is difficult to compare study results since pretreatment, substrate, vegetation, and climatic conditions vary widely among the sites. The longest study period among those cited in Table 1 was 16 months after 6 months of operation (Cathcart et al., 1994), so long term results are still lacking. Excel- lent percent reductions have been noted at constructed wetlands treating swine wastewater in Al- abama (McCaskey et al., 1994), North Carolina (Hunt et al., 1994) and Mississippi (Cathcart et al., 1994), perhaps because loading rates have tended to be lower at swine operations than at dairy farms.

In England, a 250 m 2 horizontal subsurface flow reed bed which initially received only dairy parlor washings decreased BOD by 17% and total suspended solids (TSS) by 57% (Biddlestone et al., 1991). After 6 months, barnyard runoff was added to the waste and influent BOD and TSS concentrations more than doubled. The reed bed became more effective and it reduced BOD by 49% and suspended solids by 70%. The enhanced retention despite the added waste may have been because the reeds had become better established or because the first mea- surements were taken during the winter while the second set was taken from June to September. The increased temperatures of the growing season may have enhanced BOD removal. After 2 years of operation, a greater reduction in BOD was considered

desirable, so two vertical flow reed beds were added upstream of the original bed (Biddlestone et al., 1994). In vertical flow reed beds, the wastewater flows downward through the sand and gravel matrix below the surface. Perforated pipes pass through the substrate, keeping the subsurface zone oxygenated. After the installation of the two vertical flow beds, the entire system reduced BOD concentration by 94% (from 1190 to 70 mg l - l ) .

Ammonia (NH 3) is a major constituent of dairy parlor wastewater with concentrations as high as 220 mg NH3-N 1 - l (Cronk et al., 1994). In Georgia, ammonia loading from dairy lagoons to streams was reduced by 97% after wetlands were constructed to receive the lagoon effluent (GSWCC, 1992). The pathway of removal was not clear and a portion of the nitrogen may have remained in the wetlands' outflow in the form of nitrate, which was not measured. At a dairy farm in Maryland, reduction of ammonia has been low (0-30%), perhaps due to a lack of oxygen for nitrification. Retention of BOD, TSS, total Kjeldahl nitrogen (TKN), NH3, and total phosphorus (TP) was extremely variable at the Mary- land site, probably because different management strategies were used throughout the first year of operation. Nitrate-nitrogen (NO 3) removal was an exception and influent nitrate concentrations of l0 mg NO3-N l- ~ were reduced by 90-99% throughout the study. Loadings of other parameters were consis- tently much greater, with TKN and TP concentrations often greater than 200 mg l-~ and BOD levels up to 10 000 mg l - i (Cronk et al., 1994).

In Oregon, a dairy wastewater treatment wetland has decreased influent BOD, TSS, TKN, NH 3, NO 3, TP and ortho-P by more than half in the first year of operation (Skarda et al., 1994). The researchers at- tribute ammonia decreases to volatilization rather than nitrification and denitrification because oxygen levels were too low for nitrification to occur. Volatilization may be an important ammonia removal pathway in other animal wastewater treatment wetlands as well, since BOD levels are very high and the oxidation of nitrogen is probably oxygen limited. In addition, volatilization of ammonia increases with increasing pH and pH levels in wastewater treatment wetlands are generally neutral to alkaline.

An Indiana dairy wetland reduced BOD, TSS, TKN, NH 3, NO3, TP and ortho-P by approximately


half in 1993 (Reaves et al., 1994b), however the wetland may not continue to function at this level because solids were not removed from upstream holding areas. The wetland cells received approximately 10 cm of manure in the upper third of each cell. Without more thorough pretreatment, the researchers expect the constructed wetlands to have a short effective life span.

6. Pretreatment and post-treatment

Wetlands for animal wastewater treatment should always be coupled with additional waste management strategies. Several pretreatment strategies are in use and their effectiveness is of utmost importance to the constructed wetlands' performance. The accumulation of solids shortens the effective life of a constructed wetland, making solids removal a necessary pretreatment step. Upstream settling basins, lagoons, or septic tanks can remove solids and ideally release only liquid effluent for treatment within the wetlands. For subsurface flow wetlands, Biddlestone et al. (1991) recommend a mechanical aeration device and a sedimentation area before the liquid waste flows into the bed.

On a dairy farm in Sonoma County, California, the waste from the milking parlor, loafing areas, and feeding area are flushed to a settling basin and then to a rotating screen separator (Lanier et al., 1991). The solids are separated and the liquid is discharged to a pond. The solids are dried on a field and then used as bedding or land applied and the wetlands receive the liquid portion of the waste after further settling has occurred in the pond. The wetlands alone have successfully reduced BOD 5 (97%), TSS (99%), TKN (96%), NH 3 (86%), and TP (93%; City of Santa Rosa, 1994).

High nutrient reductions have been achieved at the Sand Mountain Agricultural Experiment Station of Auburn University (McCaskey et al., 1994; Table 1) where swine waste is subjected to primary treatment in a two-stage lagoon system that reduces BOD 5 60% before the liquid waste enters the wetlands (McCaskey et al., 1994). In addition, the lagoon effluent is diluted with fresh water at a ratio of 2.7:1 (lagoon effluent: water) to reduce average in-

fluent ammonia concentration from 173 to 94 mg l- (McCaskey et al., 1994).

The Pontotoc Ridge-Flatwoods Branch Experi- ment Station in Mississippi incorporates both pretreatment and post-treatment of swine waste in its marsh-pond-marsh system. Upstream of the marshes are two anaerobic lagoons, and downstream are two vegetated strips which receive the wetland effluent. Although pollutant reduction efficiencies of the anaerobic lagoon have not been quantified, the post- treatment vegetated strips significantly enhance the system's overall performance. In the wetlands alone, mass retention of BOD 5 was 54% after 16 months of operation (Table 1). The wetlands and the and the vegetated strip reduced BOD 5 by 96% (Cathcart et al., 1994).

At the Oregon State University Experimental Wetland (for dairy wastewater) pond water is recycled to dilute the incoming waste so that influent concentrations of ammonia are kept at a maximum of 100 mg NH3-N l - I ; solids are maintained below 1500 mg 1-1; and BOD loading is 74 kg ha- l day-i (Skarda et al., 1994). In addition, a post-treatment detention pond is included in the system. Reductions of all parameters were above 50% in the first winter of operation (Table 1).

The Auburn and Mississippi State University sites treating swine wastewater receive effluent from primary and secondary lagoons (Hammer, 1993) and the North Carolina site (Hunt et al., 1994) includes an upstream anaerobic lagoon. Conversely, wetlands at some dairy farms have either no pretreatment (Biddlestone et al., 1991), only settling basins (Cronk et al., 1994; Reaves et al., 1994a), or rely on dilution to reduce influent concentrations (Skarda et al., 1994).

Many of the most elaborate pre- and post-treatment systems are at research and demonstration sites. It may be more difficult to plan and operate these components on a private farm since daily operation may be labor intensive and expensive. Nonetheless, upstream removal of BOD and TSS is essential to any animal wastewater treatment wetland because the concentrations are typically too high to be treated in a wetland (Surrency, personal communication). Without pretreatment, a constructed wetland is ineffective (Reaves et al., 1994b).

108 J.K. Cronk /Agriculture, Ecosystem and Environment 58 (1996) 97-114

7. Vegetation

In constructed wetlands, plants provide a substrate and a carbon source for microbes. Wetland plants oxygenate the substrate immediately adjacent to their roots and increase the aerobic portion of an other- wise anaerobic zone (Brix, 1993). In addition, plants remove nutrients from the incoming wastewater during the growing season. While plant nutrient uptake is usually not the major pathway of nitrogen and phosphorus removal, it has been credited with 16- 75% of total N removal and 12-73% removal of total P in wastewater treatment wetlands (Reddy and DeBusk, 1987).

A variety of plant species have been recommended for use in animal wastewater treatment wetlands (Table 3). Not all of the species in Table 3 may

be appropriate for every wetland. Desirable species are native, and therefore best suited to local conditions. They should have high productivity for rapid nutrient uptake, rhizome production, and coloniza- tion. They should be perennial to avoid annual plant- ings (Hammer, 1993) and they should be easily and inexpensively obtained. Finally, in order to survive, they should be able to tolerate high nutrient inputs. In particular, ammonia is often implicated in plant death in animal wastewater treatment wetlands, therefore a tolerance of high ammonia concentrations is necessary unless reliable pretreatment methods are used to control this pollutant.

In field trials, giant bulrush (Scirpus californicus) and softstem bulrush (Scirpus validus) survived ammonia concentrations greater than 200 mg NH3-N l- ~ while cattails were severely stressed by the same

Table 3 Recommended plant species to be used in wetlands for the treatment of animal wastewater

Common name Latin binomial Source

Emergents Arrowhead Sagittaria spp. NRCS, 1991; Hammer, 1993 Bulrush Scirpus spp. NRCS, 1991; Hammer, 1993 Canna lily Canna flaccida NRCS, 1991 Cattail Typha spp. NRCS, 1991; Hammer, 1993 Elephant ear Colocasia esculenta NRCS, 1991 Giant bulrush Scirpus californicus Surrency, 1993 Giant eutgrass Zizaniopsis milacea NRCS, 1991; Surrency, 1993 Iris Iris versicolor, L pseudacorus NRCS, 1991; Hammer, 1993 Maidencane Panicum hemitomon NRCS, 1991 Pickerelweed Pontederia cordata NRCS, 1991 Plantain Alisma spp. Hammer, 1993 Giant reed Phragmites australis NRCS, 1991 ; Biddlestone et ai., 1991 Rush Juncus spp., Cyperus spp. NRCS, 1991; Hammer, 1993

Fimbristylis spp., Eleocharis spp. Water chestnut Eleocharis dulcis NRCS, 1991

Submer gent s Coontail Ceratophyllum demersum Hammer, 1993 Naiad Najas spp. Hammer, 1993 Pondweed Potamogeton spp. Hammer, 1993 Water weed Elodea canadensis Hammer, 1994 Wild celery Valisneria americana Hammer, 1994

Floating Big duckweed Spirodelapunctata Koles et al., 1987 Duckweed Lemna spp. Koles et al., 1987

Rooted floating leaves American water lily Nelumbo lutea Hammer, 1993 Gentian Nymphoides spp. Hammer, 1993 Water lily Nymphea spp. Hammer, 1993

J.K. Cronk /Agriculture, Ecosystem and Environment 58 (1996) 97-114 109

high ammonia concentration (Surrency, 1993). Both giant and softstem bulrush rapidly produce rhizomes, and continue to grow past the end of the growing season for other plants. Giant cutgrass (Zizaniopsis milacea) has also been observed to survive and grow rapidly in animal wastewater treatment wetlands (Surrency, 1993).

Some researchers have used duckweed (Lemna spp.; Koles et al., 1987; Whitehead et al., 1987) and submerged species (Hammer, 1993) successfully. Duckweed is a volunteer species in a dairy farm wetland in Maryland and it has flourished despite ammonia concentrations as high as 160 mg NH3-N 1-l (Cronk et al., 1994). In a marsh-pond-meadow system, duckweed may inhibit phytoplankton oxygen production in the pond by shading the water column below. This loss of oxygen may inhibit nitrification (Cathcart et al., 1994).

NRCS (1991) discourages planting duckweed and submergents, arguing that they may need to be har- vested to effect permanent pollutant removals. Har- vesting can permanently remove the nutrients in plant biomass from the treatment system, however it is a time consuming process that may in fact be detrimental to the wetland's function. As plant litter accumulates, the substrate and carbon available for microbes increases. Since water quality improvement in wetlands is largely a function of having well developed microbial populations, removing plants probably does not improve the performance of most systems.

Giant reed (Phragmites australis) is commonly used in Europe, and NRCS (1991) includes it as one of its recommended species. However, others in the US discourage its use because it is an aggressive non-native species (Hammer, 1993). One study of the oxygenation of the root zone showed giant reed to be a poor oxygen transporter (Sikora, 1994) and since root zone oxygenation is vital to oxidation processes, giant reed may not be as desirable as other macrophyte species. European studies contradict these findings and have shown that giant reed can oxygenate the rhizosphere at a rate of 15 g 02 m -2 day- I (Armstrong et al., 1988), but an extensive root system takes 3 years to develop (Biddlestone et al., 1991) so system effectiveness may be delayed.

Plants may survive best if they are planted before waste enters the wetlands and then allowed to estab-

lish under wet conditions. Surrency (personal communication) recommends planting when the wetlands are dry and adding 2.5 cm of clean water per week until the desired depth is reached. At the end of 6 or 7 weeks, the wastewater can be added. If the plants are well established, they can withstand high ammonia concentrations. In addition, the initial flush of wastewater is diluted by the accumulated fresh water. McCaskey et al. (1994) suggest that plants should be allowed to establish for two growing seasons before wastewater is applied; however, this length of time before treatment may not be an option.

8. Soils

The site of an animal wastewater treatment wetland may be predetermined by the topography of the farm, existing structures, and land availability. Nonetheless, soils in the chosen area should be tested and deemed suitable to retain water before a wetland is constructed. Soil profiles can show soil type, particle size distribution, and allow classification of the soil. Hydraulic conductivity should be measured. If it is greater than considered desirable (under 10 -6

cm s - t ; Hammer, 1994), a clay or plastic liner may be necessary in order to retain standing water and prevent contamination of groundwater. In some in- stances, compacting the clay layer may provide suffi- cient water retention that the wetland's substrate will be sealed as hydric conditions develop. If plastic or geo-textiles are used to line the basin, a soil layer must be added to provide a plant substrate. The thicker the liner, the more expensive it is, so the addition of soil rich in bentonite may be a more practical solution.

9. Cost

A number of questions regarding cost arise before a wetland is built. The amount of funding available, and the limits and rules concerning its expenditure are important considerations in the planning stages of a constructed wetland (Newling, 1982). An estimate of a new wetland's cost should include (Toml- janovich and Perez, 1989): 1. engineering plan; 2. preconstruction site preparation;


3. construction (labor, equipment, materials, super- vision);

4. long-term management; 5. pest control; 6. contingency costs.

The cost of construction depends on location, type, treatment objectives, and the maintenance required. Factors that can add to the cost are access to the site, substrate characteristics, the addition of a clay or plastic liner, cost of protective structures, local labor rates, and the availability of equipment (Newling, 1982).

Much of the cost information available for animal wastewater treatment wetlands is for research and demonstration sites rather than wetlands on private farms (Table 4). Research sites tend to include a greater number of cells and structures for experiments and therefore the cost of these is probably higher than on private farms. The average cost of construction per cell is $2980 (1988-1993 US dol- lars). If a farmer constructs only one large cell, the total cost may be lower because labor and plumbing structures would decrease. Hammer et al. (1991) estimated that farmers providing their own labor and doing minimal planting could keep the total cost of a treatment wetland at $3000 US.

While constructed wetlands may be an inexpensive alternative for waste management, the cost of maintenance must be added to the construction cost. Maintenance includes active management of the solid portion of the waste. The amount of solids entering the wetland must be minimized in order for the wetland to effectively treat the waste. Solids man-

agement is an ongoing problem for the farmer and this duty cannot be ignored once wetlands have been installed (Healy and McLoud, 1994; Reaves et al., 1994b). In addition, water levels may need to be manipulated in order to maintain wet conditions (Holmes et al., 1994) and the surrounding area may need to be cleared of vegetation periodically in order to allow access to the wetlands. A work plan and a budget for daily operation should be included in the installation of a wetland to treat animal waste (Bal- dwin and Davenport, 1994; Healy and McLoud, 1994). Further, some wetlands fail to operate as originally designed and constructed. Farm managers must be willing to adopt a flexible, iterative approach to develop an optimum design over a period of years. Thus, all costs cannot be anticipated.

I0. Conclusions and recommendations

Properly designed, constructed, and maintained animal wastewater treatment wetlands can effectively reduce or eliminate contaminant loads to downstream waters. Constructed wetlands are relatively inexpensive and easy to construct and maintain. These attributes should be emphasized if wetlands are to be adopted as a best management practice (BMP) by private landowners. A wetlands BMP should be only one component of a well integrated farm waste management strategy that must include pretreatment and adequate temporary waste storage structures. With current knowledge, animal wastewater treatment wetlands can be recommended under

Table 4 Cost of construction of dairy and swine wastewater treatment wetlands ($US)

Site Cost of Cost m-2 No. of Cost per Year of Source construction cells c e l l construction

Dairy Franklin Co., AL 11490 12.76 5 2298 1989 Hammer and Watson, 1988 Newton Co., MI 19079 93.07 6 3180 1989 NRCS, 1989 Calumet Co., WI 39600 250.63 12 3300 1992 Holmes et al., 1992 Frederick Co., MD 9000 8.24 2 4500 1993 Baldwin and Davenport, 1994 Kent Co., MD 7000 5.83 2 3500 1993 Baldwin and Davenport, 1994 a

Swine Auburn Univ., AL 12800 2.25 10 1280 1988 Hammer, 1993 Worcester Co., MD 20000 1.45 6 3333 1993 Baldwin and Davenport, 1994

a Estimated cost is given. Actual cost to the farmer was zero because all labor, materials and equipment were donated.

J.K. Cronk / Agriculture, Ecosystem and Environment 58 (t996) 97-114 111

circumstances favorable to their operation: high moisture and available land area.

However, constructed wetlands may not be an ideal best management practice on every farm. A definitive design does not exist, so wetlands may have to be altered after construction and the success of any given design is not guaranteed. Adjustments may have to be made to correct construction problems or to compensate for management changes (Holmes et al., 1994; Reaves et al., 1994b). Con- structed wetlands may not effectively remove pollu- tants immediately after construction. A period of time in which the plant and microbial communities become established may be necessary before removal rates reach acceptable levels.

mined (Tchobanoglous, 1993). Designs based on well-defined constituents and their decay rates may result in better waste treatment.

11.3. What are the maximum concentrations of nutrients and BOD that can be treated in a wetland?

Is it unrealistic to expect that wetlands can treat waste loads above a certain level? What loading is too high? Should the loading rate recommendations vary under different climatic or soil conditions? Do the current design equations adequately account for extremely high loads? Mesocosm experiments with a series of loading rates would help answer these questions.

11. Future research needs 11.4. Which pretreatment structures are most effective?

Results from research currently underway will improve our ability to design effective wetlands, but a number of questions concerning animal wastewater treatment wetlands remain. Future research must answer the following questions:

11.1. Are current design guidelines adequate?

Many of the animal wastewater treatment wetlands are designed according to the guidelines of the NRCS (1991). Do these procedures result in the best design for treating animal wastewater? If so, under what circumstances does it function at a maximum? Can adjustments to the guidelines be made that will improve wetland performance? With present data, it is difficult to determine whether any given design is preferable. Results are not easily comparable since wetlands of various designs treat different wastes, use different pretreatment structures, and are in different areas of the U.S. Different design methods could be used to construct a number of wetlands at a single site, to test their effectiveness in treating the same wastewater under the same conditions.

11.2. Should the design be based on BOD loadings or on other parameters?

The major drawback of basing the design on BOD is that the source of the oxygen demand is undeter-

A variety of pretreatment structures have been built in conjunction with wetlands, but very few data show the effectiveness of these devices. What structures are the most effective in reducing the solid and BOD load to wetlands? Is aeration of waste to enhance nitrification feasible? Different pretreatment strategies should be tested at a single site to discern which practice best enhances wetland performance. Recommendations for pretreatment should go hand in hand with wetland construction.

11.5. Are site specific results applicable to other areas?

What climatic conditions must be met in order for animal wastewater treatment wetlands to be effective? Should designs be different for different re- gions?

11.6. How can the cost to the farmer be minimized?

Research concerning the use of constructed wetlands should focus on maximizing their potential for contaminant removal while minimizing their cost.

The goal of research on all animal wastewater treatment structures, including wetlands, is to foster the widespread treatment of animal wastewater in order to protect the quality of surface and ground water. Therefore it is necessary to determine under


what circumstances wetlands can be added to an animal farmer's waste management choices and how they can be designed to function at an optimum.

Acknowledgements

Herbert Brodie, Hugh Crowell, Tracy Davenport, William Magette, and two anonymous reviewers provided useful comments on earlier drafts of this manuscript. Patricia Lupo typed Table 2. Lovant Hicks drew the figures. This work was partially supported by a grant from the US EPA and by a University of Maryland Graduate Research Council Summer Grant.

References

Armstrong, J., Armstrong, W. and Beckett, P.M., 1988. Phrag- mites australis: a critical appraisal of the ventilating pressure concept and an analysis of resistance to pressurized gas flow and gaseous diffusion in horizontal rhizomes. New Phytol., 110: 383-389.

Baldwin, A.P. and Davenport, T.N., 1994. Constructed wetlands for animal waste treatment: a progress report of three case studies in Maryland. In: P.J. DuBowy and R.P. Reaves (Edi- tors), Proceedings of a Workshop on Constructed Wetlands for Animal Waste Management, 4 -6 April 1994, Lafayette, IN, pp. 103-117.

Biddlestone, A.J., Gray, K.R. and Thurairajan, K., 1991. A botan- ical approach to the treatment of wastewaters. J. Biotechnol., 17: 209-220.

Biddlestone, A.J., Gray, K.R. and Thayanithy, K., 1994. Compost- ing reed beds for aerobic treatment of livestock wastes. In: I.A. Dewi, R.F.E. Axford, I. Fayez, M. Marai and H.M. Oured (Editors), Pollution in Livestock Production Systems. CAB International, Wallingford, UK, pp. 345-360.

Brix, H., 1993. Macrophyte-mediated oxygen transfer in wetlands: transport mechanisms and rates. In: G.A. Moshiri (Editor), Constructed Wetlands for Water Quality Improvement. Lewis Publishers, Boca Raton, FL, pp. 391-398.

Brodie, H.L., 1988. Estimated distribution of farm animals and their manure nutrient production in Maryland by watershed-- 1988. Agricultural Engineering Department, University of Maryland.

Brodrick, S.J., Cullen, P. and Maher, W., 1988. Denitriflcation in a natural wetland receiving secondary treated effluent. Water Resourc., 22: 431-439.

Cathcart, T.P., Hammer, D.A. and Triyono, S., 1994. Performance of a constructed wetland-vegetated strip system used for swine waste treatment. In: P.J. DuBowy and R.P. Reaves (Editors), Proceedings of a Workshop on Constructed Wet-

lands for Animal Waste Management, 4 -6 April 19941 Lafayette, IN, pp. 9-22.

Chen, S., Cothren, G.M., DeRamus, H.A., Langlinais, S., Huner, J.V. and Malone, R.F., 1995. Design of constructed wetlands for dairy waste water treatment in Louisiana. In: K. Steele (Editor), Animal Waste and the Land-Water Interface. Lewis Publishers, Boca Raton, FL, pp. 197-204.

City of Santa Rosa, 1994. Lafranchi Dairy waste treatment marsh. Unpublished report, 10 pp.

Cronk, J.K., Shirmohammadi, A. and Kodmur, V., 1994. An evaluation of wetlands for the treatment of dairy effluent: results from the first year of operations. Presented at the International Winter Meeting, Pap. No. 94-2600, ASAE, St. Joseph, MI, USA.

Dierberg, F.E. and Brezonik, P.L., 1983. Nitrogen and phosphorus mass balances in natural and sewage-enriched cypress domes. J. Appl. Ecol., 20: 323-337.

Faulkner, S.P. and Richardson, C.J., 1989. Physical and chemical characteristics of freshwater wetland soils. In: D.A. Hammer (Editor), Constructed Wetlands for Wastewater Treatment. Lewis Publishers, Chelsea, MI, pp. 41-72.

Gambrell, R.P. and Patrick, W.H., Jr., 1978. Chemical and micro- biological properties of anaerobic soils and sediments. In: D.D. Hook and R.M. Crawford (Editors), Plant Life in Anaer- obic Environments. Ann Arbor Science Publishers, Ann Ar- bor, MI, pp. 375-423.

Gearhart, R.A., Klopp, F. and Allen, G., 1989. Constructed free surface wetlands to treat and receive wastewater: pilot project to full scale. In: D.A. Hammer (Editor), Constructed Wetlands for Wastewater Treatment. Lewis Publishers, Chelsea, MI, pp. 121-138.

GSWCC, 1992. Constructed wetlands assimilation of effluent from a confined swine operation. Technical Report, Georgia Soil and Water Conservation Commission.

Gersberg, R.M., Gearhart, R.A. and Ives, M., 1989. Pathogen removal in constructed wetlands. In: D.A. Hammer (Editor), Constructed Wetlands for Wastewater Treatment. Lewis Pub- lishers, Chelsea, MI, pp. 431-445.

Hammer, D.A. (Editor), 1989. Constructed Wetlands for Wastew- ater Treatment. Lewis Publishers, Chelsea, MI, 831 pp.

Hammer, D.A., 1992. Designing constructed wetland systems to treat agricultural nonpoint source pollution. Ecol. Eng., 1: 49-82.

Hammer, D.A., 1993. Constructed wetlands for wastewater treatment: an overview of a low cost technology. Presented at the Constructed Wetlands for Wastewater Treatment Conference, Middleton, County Cork, Ireland.

Hammer, D.A., 1994. Guidelines for design, construction and operation of constructed wetlands for livestock wastewater treatment. In: P.J. DuBowy and R.P. Reaves (Editors), Pro- ceedings of a Workshop on Constructed Wetlands for Animal Waste Management, 4 -6 April 1994, Lafayette, IN, pp. 155- 181.

Hammer, D.A. and Watson, J.T., 1988. Agricultural waste treatment with constructed wetlands. In: Proc. of the Symp. on the Protection of Wetlands from Agricultural Impacts, 25-29 April 1988, Fort Collins, CO, pp. 161-165.


Hammer, D.A., Pullin, B.P., McCaskey, T.A., Eason, J. and Payne, V.W.E., 1991. Treating livestock wastewaters with constructed wetlands. Constructed Wetlands for Water Quality Improvement Conference, Pensacola, FL.

Hammer, D.A. and Knight, R.L., 1994. Designing constructed wetlands for nitrogen removal. Water Sci. Technol., 29: 15-27.

Healy, J.W. and McLoud, P.R., 1994. Use of constructed wetlands and infiltration areas in NRCS approved waste management systems. In: P.J. DuBowy and R.P. Reaves (Editors), Proceed- ings of a Workshop on Constructed Wetlands for Animal Waste Management, 4-6 April 1994, Lafayette, IN, pp. 65-71.

Holmes, B.J., Massie, L.R., Bubenzer, G.D. and Hines, G., 1992. Design and construction of a wetland to treat milkhouse wastewater. Presented at the International Summer Meeting, Pap. No. 92-4524, ASAE, St. Joseph, MI, USA.

Holmes, B.J., Bubenzer, G.D., Massie, L.R. and Hines, G., 1994. A constructed wetland for treating milkhouse wastewater in a cold climate--status report. In: P.J. DuBowy and R.P. Reaves (Editors), Proceedings of a Workshop on Constructed Wet- lands for Animal Waste Management, 4-6 April 1994, Lafayette, IN, pp. 53-64.

Holmes, B.J., Doll, B.J., Rock, C.A., Bubenzer, G.D., Kostinec, R. and Massie, L.R., 1995. Experiences with two constructed wetlands for treating milking center waste water in a cold climate. In: K. Steele (Editor) Animal Waste and the Land- Water Interface. Lewis Publishers, Boca Raton, FL, pp. 223- 230.

Hunt, P.G., Humenik, F.J., Szogi, A.A., Rice, J.M., Stone, K.C. and Sadler, E.J., 1994. Swine wastewater treatment in constructed wetlands. In: Proc. of the 2nd Conf. on Environmen- tally Sound Agriculture, University of Florida, pp. 268-275.

Hunt, P.G., Stone, K.C., Humenik, F.J. and Rice, J.M., 1995. Impact of animal waste on water quality in an eastern coastal plain watershed. In: K. Steele (Editor), Animal Waste and the Land-Water Interface. Lewis Publishers, Boca Raton, FL, pp. 257-264.

James, B.R., Rabenhorst, M.C. and Frigon, G.A., 1992. Phospho- rus sorption by peat and sand amended with iron oxides or steel wool. Water Environ. Res., 64(5): 699-705.

Jaworski, N.A., Groffman, P.M., Keller, A.A. and Prager, J.C., 1992. A watershed nitrogen and phosphorus balance: the Upper Potomac River Basin. Estuaries, 15: 83-95.

Johnston, C.A., 1991. Sediment and nutrient retention by freshwater wetlands: effects on surface water quality. Crit. Rev. Environ. Control, 21: 491-565.

Johnston, C.A., Bubenzer, G.D., Lee, G.B., Madison, F.W. and McHenry, J.R., 1984. Nutrient trapping by sediment deposition in a seasonally flooded lakeside wetland. J. Environ. Qual., 13: 283-290.

Kadlec, R.H., 1985. Aging phenomenon in wastewater wetlands. In: P.J. Godfrey (Editor), Ecological Considerations in Wet- land Treatment of Municipal Wastewaters. Van Nostrand Reinhold, New York, pp. 338-350.

Kadlec, R.H., 1987. Northern natural wetland water treatment systems. In: K.R. Reddy and W.H. Smith (Editors), Aquatic

Plants for Water Treatment and Resource Recovery. Magnolia Publishing, Orlando, FL, pp. 83-98.

Kashmanian, R.M., 1994. Building support for composting in agriculture. Biocycle, 35(12): 67-70.

Knight, R.L., McKim, T.W. and Kohl, H.R., 1987. Performance of a natural wetland treatment system for wastewater management. J. Water Pollut. Control Fed., 59: 746-754.

Koles, S.M., Petrell, R.J. and Bagnall, L.O., 1987. Duckweed culture for reduction of ammonia, phosphorus and suspended solids from algal-rich water. In: K.R. Reddy and W.H. Smith (Editors), Aquatic Plants for Water Treatment and Resource Recovery. Magnolia Publishing, Orlando, FL, pp. 769-774.

Krider, J.N. and Boyd, W.H., 1992. NRCS technical requirements for constructed wetlands for agricultural wastewater treatment. Presented at the Int. Winter Meeting, Pap. No. 92-4523, ASAE, St. Joseph, MI, USA.

Lanier, A.L., Fox, D. and Smith, D.W., 1991. Wetland for treating liquid dairy waste design and monitoring. Presented at the Int. Summer Meeting, Pap. No. 91-4020, ASAE, St. Joseph, MI, USA.

McCaskey, T.A., Britt, S.N., Hanhah, T.C., Eason, J.T., Payne, V.W.E. and Hammer, D.A., 1994. Treatment of swine lagoon effluent by constructed wetlands operated at three loading rates. In: P.J. DuBowy and R.P. Reaves (Editors), Proc. of a Workshop on Constructed Wetlands for Animal Waste Man- agement, 4-6 April 1994, Lafayette, IN, pp. 23-33.

Midwest Plan Service, 1987. Structures and Environment Hand- book, MWPS-I 1 lth edn. Iowa State University, Ames.

Mitsch, W.J. and Cronk, J.K., 1992. Creation and restoration of wetlands: some design considerations for ecological engineering. Adv. Soil Sci., 17: 217-259.

Mitsch, W.J. and Gosselink, J.G., 1993. Wetlands, 2nd edn. Van Nostrand Reinhold Co., New York, 722 pp.

NRCS, 1989. Project Description: Use of Constructed Wetlands to Treat Wastewater from Confined Animal Operations in Mis- sissippi. Natural Resource Conservation Service, United States Department of Agriculture, Washington, DC. (Unpublished.)

NRCS, 1991. Technical requirements for agricultural wastewater treatment. Natl. Bull. No. 210-1-17, Natural Resource Conser- vation Service, United States Department of Agriculture, Washington, DC.

NRCS, 1992. Agricultural Waste Management Field Handbook, Part 651. National Engineering Handbook 210-VI. Natural Resource Conservation Service, United States Department of Agriculture, Washington, DC.

Newling, C.J., 1982. Feasibility report on a Santa Ann River marsh restoration and habitat development project. In: M.C. Landin (Editor), Habitat Development at Eight Corps of Engi- neers Sites: Feasibility and Assessment. US Army Engineer Waterways Experiment Station Environmental Laboratory, Vicksburg, MS, pp. 45-84.

Odum, H.T., Ewel, K.C., Mitsch, W.J. and Ordway, J.W., t977. Recycling treated sewage through cypress wetlands in Florida. In: F.M. D'Itri (Editor), Wastewater Renovation and Reuse. Marcel Dekker, New York, pp. 35-67.


Rastorfer, D.D. and Schneider, J.H., 1994. Lincoln University constructed wetland for swine lagoon effluent treatment. Pre- sented at the 1994 Internatinnal Winter Meeting of the Ameri- can Society of Agricultural Engineers in Atlanta, Georgia. Pap. No. 942599.

Reaves, R.P., DuBowy, P.J., Jones, D.D. and Sutton, A.L., 1994a. Design of an experimental constructed wetland for treatment of swine lagoon effluent. In: P.J. DuBowy and R.P. Reaves (Editors), Proc. of a Workshop on Constructed Wetlands for Animal Waste Management, 4-6 April 1994, Lafayette, IN, pp. 82-96.

Reaves, R.P., DuBowy, P.J. and Miller, B.K., 1994b. Performance of a constructed wetland for dairy waste treatment in Lagrange County, Indiana. In: P.J. DuBowy and R.P. Reaves (Editors), Proc. of a Workshop on Constructed Wetlands for Animal Waste Management, 4-6 April 1994, Lafayette, IN, pp. 43-52.

Reddy, K.R. and DeBusk, W.F., 1987. Nutrient storage capabili- ties of aquatic and wetland plants. In: K.R. Reddy and W.H. Smith (Editors), Aquatic Plants for Water Treatment and Resource Recovery. Magnolia Publishing, Orlando, FL, pp. 337-357.

Reed, C.S., Middlebrooks, J.E. and Crites, W.R., 1988. Natural Systems for Waste Management and Treatment. McGraw-Hill, New York, pp. 166-183.

Richardson, C.J., 1985. Mechanisms controlling phosphorus retention capacity in freshwater wetlands. Science, 228:1424-1427.

Schofleld, K., Seager, J. and Merriman, R.P., 1990. The impact of intensive dairy farming activities on river quality: the Eastern Cleddau catchment study. J. IWEM, 45: 176-186.

Sikora, F.J., 1994. Summary of research at the Tennessee Valley Authority's constructed wetlands research and development facility. In: P.J. DuBowy and R.P. Reaves (Editors), Proc. of a

Workshop on Constructed Wetlands for Animal Waste Man- agement, 4--6 April 1994, Lafayette, IN, pp. 1-18.

Skarda, S.M., Moore, J.A., Niswander, S.F. and Gamroth, M.J., 1994. Preliminary results of wetland for treatment of dairy farm wastewater. In: P.J. DuBowy and R.P. Reaves (Editors), Proc. of a Workshop on Constructed Wetlands for Animal Waste Management, 4-6 April 1994, Lafayette, IN, pp. 34-42.

Surrency, D., 1993. Evaluation of aquatic plants for constructed wetlands. In: G.A. Moshiri (Editor), Constructed Wetlands for Water Quality Improvement. Lewis Publishers, Boca Raton, FL, pp. 349-357.

Szogi, A.A., Hunt, P.G., Humenik, F.J., Stone, K.C., Rice, J.M. and Sadler, E.J., 1994. Seasonal dynamics of nutrients and physico-chemical conditions in a constructed wetland for swine wastewater treatment. Presented at the Int. Winter Meeting, Pap. No. 94-2602, ASAE, St. Joseph, MI, USA.

Tchobanoglous, G., 1993. Constructed wetlands and aquatic plant systems: research, design, operational, and monitoring issues. In: G.A. Moshiri (Editor), Constructed Wetlands for Water Quality Improvement. Lewis Publishers, Boca Raton, FL, pp. 23-34.

Tomljanovich, D.A. and Perez, O., 1989. Constructing the wastewater treatment wetland--some factors to consider. In: D.A. Hammer (Editor), Constructed Wetlands for Wastewater Treatment. Lewis Publishers, Chelsea, MI, pp. 399-404.

Whitehead, A.J., Lo, K.V. and Bulley, N.R., 1987. The effect of hydraulic retention time and duckweed cropping rate on nutrient removal from dairy barn wastewater. In: K.R. Reddy and W.H. Smith (Editors), Aquatic Plants for Water Treatment and Resource Recovery. Magnolia Publishing, Orlando, FL, pp. 697-702.

Documents

Constructed wetlands to treat wastewater from dairy and ... wetlands to tre… · Constructed wetlands to treat wastewater from dairy ... are removed from the water by plant uptake