Nitrogen Source Effects on Ammonia Volatilization

as Measured with Semi-Static Chambers

Claudia Pozzi Jantalia, Ardell D. Halvorson,* Ronald F. Follett,

Agronomy, Soils & Environmental Quality

Bruno Jose Rodrigues Alves, Jose Carlos Polidoro, and Segundo Urquiaga

ABSTRACT
Ammonia (NH3) volatilization is one of the main pathways of N loss from agricultural cropping systems. This study evaluated
the NH3 N loss from four urea-based N sources (urea, ureaammonium nitrate [UAN], SuperU, and ESN [polymer-coated
urea]) surface band applied at a rate of 200 kg N ha1 to irrigated, strip-till corn production systems for 2 yr using semi-static
chambers (semi-open and open) to measure NH3 N loss. The efficiency of the semi-static chambers in estimating NH3 N loss
under field conditions was determined using 15N labeled urea applied at rates of 50, 100, and 200 kg N ha1. Both chamber types
had similar NH3 N recoveries and calibration factors. Immediate irrigation with 16 to 19 mm of water 1 d after N fertilization
probably limited NH3 N volatilization from surface-applied N fertilizers to a range of 0.1 to 4.0% of total N applied. SuperU,
which contains a urease inhibitor, had the lowest level of NH3 N loss when compared to the other N sources. Analyzed across
years, estimated NH3 N losses for the N sources were in the order: ESN = UAN > urea > SuperU. Both years the results showed
that measurement time may need to be increased to evaluate NH3 N volatilization from polymer-coated urea N sources such
as ESN. The open-chamber method was a viable, low cost method for estimating NH3 N loss from small field plot N studies.

N itrogen fertilizer consumption has increased

more rapidly than that of P or K to support world food
supplies (FAO, 2006). Improperly managed, N fertilizers can
form gaseous or soluble compounds with potential to pollute
air (NH3, NOx, N2O) or contaminate surface and ground
waters with nitrate N (Aneja et al., 2003). Ammonia is the
most reduced form of reactive N in the atmosphere and is
increasing from human activities (Aneja et al., 2008) with an
unpredictable fate of when and where it will be redeposited
into the global N cycle (Galloway et al., 2004). Agriculture
represents 20 to 80% of NH3N emissions in many countries
with livestock manures and N fertilizers as major sources (Mis-
selbrook et al., 2000; Aneja et al., 2008; Zhang et al., 2010).
Ammonia volatilization is one of the main pathways that N is
lost from organic and inorganic N fertilizer application (Ma et al.,
2010). This can result in low nitrogen use efficiencies (NUE) by
crops (Bouwman et al., 2002; Rochette et al., 2008). The amount
of NH3 volatilized from surface-applied N fertilizer is controlled
by many interacting soil characteristics and climatic factors and
is difficult to predict (McGinn and Janzen, 1998). Over the past
35 to 40 yr, development of methodology to measure NH3 losses
C.P. Jantalia, B.J.R. Alves, and S. Urquiaga, Embrapa Agrobiology, Road BR
Km 7, Seropedica, Rio de Janeiro 23890-000, Brazil; A.D. Halvorson, and R.F.
Follett, USDA-ARS, 2150 Centre Ave, Bldg. D, Ste. 100, Fort Collins, CO
80526; and J.C. Polidoro, Embrapa Soil, Rio de Janeiro, 22460-000, Brazil.
Trade names and company names are included for the benefit of the reader
and do not imply any endorsement or preferential treatment of the product by
the authors or the USDA-ARS. USDA is an equal opportunity provider and
employer. Contribution from USDA-ARS, Fort Collins, CO. Received 7 June
2012. *Corresponding author (ardell.halvorson@ars.usda.gov).
Published in Agron. J. 104:15951603 (2012)
Posted online 5 Sep 2012.
doi:10.2134/agronj2012.0210
Copyright 2012 by the American Society of Agronomy, 5585 Guilford
Road, Madison, WI 53711. All rights reserved. No part of this periodical may
be reproduced or transmitted in any form or by any means, electronic or
mechanical, including photocopying, recording, or any information storage
and retrieval system, without permission in writing from the publisher.
under field condition has led to an improved understanding of
major factors that drive this loss under dynamic field conditions
(McGinn and Janzen, 1998, Misselbrook and Hansen, 2001).
Conditions that favor gaseous loss of NH3 from surface-applied
N includes high crop residue, warm temperature (>13C), a drying
soil surface (water vapor loss from surface), neutral or alkaline
soil pH, and low cation exchange capacity (as found in sandy soil)
(Clay et al., 1990; Ferguson and Kissel, 1986; Bouwmeester et al.,
1985; Sommer and Christensen, 1992).
Advances in N management and fertilizer technology have
made it possible to reduce NH3N losses (Snyder, 2008).
Enhanced-efficiency N fertilizers that control N release have
been available in the U.S. fertilizer market for several years,
but their use has been limited due to their higher cost (Stewart,
2008). However, increasing N fertilizer prices, heightened
environmental awareness, increasing area under conservation
tillage agriculture and improved manufacturing technology
has led to increased interest in enhanced-efficiency N fertilizers
(Rochette et al., 2009; Halvorson et al., 2010).
Different techniques have been employed to measure NH3 loss
from N sources applied to soils (Misselbrook et al., 2005; McGinn
and Janzen, 1998; Cabrera et al., 2001; Pacholski et al., 2006).
Most of these measurement techniques represent major categories:
micrometeorological, enclosure, and indirect methods. The
micrometeorological (direct) and 15N balance (indirect) methods
(McGinn and Janzen, 1998) allow absolute estimates of NH3N
loss in the field because they do not interfere with atmospheric
conditions near the soil surface or affect NH3 volatilization.
Evaluating the effects of N source, N application rate, or
soil management on NH3 emissions in field environments
generally requires several treatments and replications, and small
plots (<50 m2) where the fetch criterion is not adequate to
Abbreviations: ESN, polymer-coated urea produced by Agrium Advanced
Technologies, Loveland, CO; N1, 50 kg N ha1; N2, 100 kg N ha1; N3,
200kg N ha1; NUE, nitrogen use efficiency; UAN, ureaammonium nitrate.

A g ro n o my J o u r n a l Vo l u m e 10 4 , I s s u e 6 2 012 1595
employ micrometeorological techniques to measure NH3N
volatilization losses (Misselbrook et al., 2005; McGinn and
Janzen, 1998). Even though there are several enclosure methods
to measure NH3N loss, static chambers are quite often used
because they permit monitoring multiple treatments in the
same crop season, have a low cost, and require reagents and
materials commonly available in laboratories worldwide (Grant
et al., 1996; Ma et al., 2010). The most used static chamber
was developed by Nmmik (1973) and was based on a semi-
open design. It is composed of a cylinder anchor inserted into
the soil connected with a chamber containing two polyfoam
sponges soaked in acid solution that act as an NH3N gas
trap. Ammonia gas reacts readily with the acidic medium with
the most common options being phosphoric (Nmmik, 1973;
Cantarella et al., 2008), sulfuric (Sherlock et al., 1989; Lara-
Cabezas and Trivelin, 1990), or hydrochloric acid (Janzen and
McGinn, 1991).
Recently a simple, improved, and inexpensive semi-static open
chamber with free air movement inside was reported by Arajo et
al. (2009). The goals for this apparatus were to have no soil anchor
requirement and acid NH3N trap placed inside the chamber to
permit free-air circulation. To improve the confidence in these
chamber techniques, previous authors have reported on indirect
15N balance measurement for both NH N recovery efficiency
3
and chamber effects on NH3 volatilization (Nmmik, 1973;
Cabezas et al., 1999; Arajo et al., 2009).
The objective of this study was to evaluate NH3N volatilization
losses from four urea-based N fertilizers applied to small field plots
(Halvorson et al., 2011) using two existing semi-static chambers
(Nmmik [1973] semi-open chamber without soil anchors
and Arajo et al. [2009] open chamber) following N fertilizer
application to a soil under irrigated corn (Zea mays L.) production.
MATERIALS AND METHODS
The study was conducted at the Agricultural Research
Development and Education Center (ARDEC) near Fort
Collins, CO (1555 m above sea level; 40396 N, 1045957
W) in the same plot area described by Halvorson et al. (2011).
The region has a semiarid temperate climate, with a typical
mean temperature of 10.6C and rainfall of 383 mm yr1
(average for 19002005). The soil was a Fort Collins clay loam
(fine-loamy, mixed, superactive, mesic Aridic Haplustalf)
with pH of 7.8 (soil: 0.001 M CaCl2 solution ratio 1:2). Other
selected soil properties were an electrical conductivity of 0.34
to 0.81 mS cm1 (soil/water ratio 1:1); soil organic C of 12 g
kg1; and total soil N of 1.28 g kg1 (Halvorson et al., 2010,
2011). The soil texture was 409 g kg1 of sand, 254 g kg1 of silt
and 337 g kg1 of clay content (Halvorson et al., 2010).
Two experiments were conducted: (i) use of 15N labeled urea
to calibrate the efficiency of the semi-static chambers (semi-
open and open) for measuring total NH3N loss from three N
rates (50, 100, and 200 kg N ha1) under our field conditions
and (ii) use of the semi-static chambers to measure NH3N
loss from four inorganic fertilizer N sources that were surface
band applied to small field plots in 2010 and 2011.
Chamber Description
The Semi-Open Chamber. The chamber evaluated in this
study was based on a semi-open model described by Nmmik
1596 Agronomy Journal Volume 104, Issue 6 2012
(1973) with the main variation in our study being the absence
of a soil anchor. The semi-open chamber was a rigid polyvinyl
chloride (PVC) cylinder, 200 mm in length by 150 mm diam.
(Fig. 1). A rigid plastic cap, 210 mm diam., was positioned
approximately 70 mm above the top of the cylinder to shed
rainfall or irrigation water. Inside each unit were fitted two
polyethylene foam (polyfoam) disc pieces 25 mm thick and 160
mm in diameter that were soaked in acid solution (40 mL) as
described below and carried from the laboratory to the field site
in air-tight plastic bags. These discs were tightly fitted within
the cylinder to reduce air exchanges with outside air. The upper
disc was inserted within 20 mm below the top of the cylinder to
reduce atmospheric NH3 contamination and replaced after 6 to
7 d. The bottom disc was located between the soil and top disc
and inserted 50 mm above the soil surface to trap NH3N losses
from soil and fertilizer and was replaced on each sampling day.
Before inserting the foam disc, the excess acid was removed by
squeezing the disc, leaving ~35 mL of acid solution remaining in
the disc before the chamber was placed on the soil surface.
The Open Chamber. This chamber was based on work
reported by Arajo et al. (2009). It consists of a transparent
polyethylene terephthalate bottle (PET, soda bottle 2 L)
with bottom removed (Fig. 2). The chamber was 260 mm in
length and 100 mm diam. The cut off bottle bottom piece
was positioned with wires approximately 20 mm above the
top opening to shed rainfall or irrigation water. Inside each
chamber was a 250 mm long wire designed with a hook to
support it from the top of the bottle, and wire basket on the
bottom end to support a plastic jar (60 mL) that contained
the acid solution to keep the foam strip moist during sampling
periods. Fastened to the top of the wire support was a strip of
polyfoam (250 mm in length, 25 mm wide, and 3 mm thick)
that had been soaked in acid solution, with the bottom end of
the foam remaining inside the plastic jar. Plastic jars were used
to carry the acid immersed foam strips from the laboratory to
experimental site for use during NH3 sampling procedures.
Sampling and Ammonia Extraction Procedures
On the day of sampling, the foam discs in the semi-open
chambers and foam strips and plastic jar in open chambers
were collected and replaced with new foam traps. To avoid
contamination of samples, disposable latex gloves worn during
the sampling activities were frequently replaced. The traps from
each chamber sampled were placed immediately in an acrylic
jar with 125 mL 2 M KCl, sealed and placed in an ice chest
maintained at 5 to 10C until their return to the laboratory.
In the laboratory, the acid solution from each sample was
transferred with a funnel to a 250 mL volumetric flask. The
acrylic jars and traps were thoroughly rinsed two more times
with 2 M KCI solution and manual shaking (10 s) to extract
the trapped ammonia from the foam strips and discs, then
brought to volume. An aliquot of total solution (~50 mL)
was filtered into plastic vials, sealed, and frozen until analysis.
Immediately after thawing, an aliquot of the solution was
analyzed for ammonia concentration using a continuous flow
analyzer QuickChem (Lachat QuickChem FIA+8000 Series,
Lachat Instruments, Loveland, CO), with Method 13-107-06-
2-D (Determination of total Kjeldahl N in soils and plants by
flow injection analysis).
 Chamber Calibration (2010)
The chamber calibration was conducted using plastic plates
to hold the soil fertilized with 3 N rates that were placed
within a check plot (no N applied for 11 yr) located in an
irrigated, strip tillage continuous corn N study (Halvorson et
al., 2011). The soil used in the chamber calibration experiment
was collected from the 0- to 10-cm depth of an unfertilized
plot area within the corn N study area. After collection, the soil
was air dried (72 h) passed through a 2-mm sieve, and mixed.
This soil was used in the chamber calibration study.
Plastic plates, 17 cm in diameter, were used to hold 400 g of the
collected air-dry soil. Two days before N fertilization, water was

Fig. 1. Schematic representation of semi-open chamber parts for NH 3 measurement.

Fig. 2. Schematic representation of open chamber parts for NH3 measurement.

Agronomy Journal Volume 104, Issue 6 2012 1597

applied to wet the soil on each plate to near field capacity. The 15N mass isotope-ratio spectrometer (model VG-903 Europa
labeled urea fertilizer was applied at the desired N rate to each Scientific Ltd., Crewe, England).
plate of soil as a liquid urea solution (Table 1). The 15N urea labeled Volatilized NH3 trapped by the chambers for each N
solution was surface band applied with a pipette throughout the treatment was expressed as NH3N loss between sampling dates
plate center (3 cm width band) on 17 June 2010. The soil plates and cumulative NH3N loss with time (mg NH3N plate1)
were immediately placed on a plastic sheet covering the soil in the calibration experiment. In the calibration experiment,
surface between the corn rows inside the field check plot area. The chamber efficiency in trapping NH3N loss was examined by
transparent plastic sheet (30 by 30 cm) was used to avoid unlabeled statistical analysis between total NH3N loss with a chamber
soil or residue contamination of the treated soil on the plates. present, as determined by 15N balance method, and NH3N
Chambers fitted with ammonia traps were immediately placed trapped by the acid traps in each chamber type. The calibration
directly on the treated soil surface. To keep the chambers upright factors for both chamber types were determined by correlation
under windy conditions, three wire stakes (5 mm diam. and 25 cm between absolute amount of NH3N loss determined in the
length) were inserted 10 cm into the soil around each plate and control treatment (without chambers) and cumulative amounts
rubbers bands (4 mm thick) were used to surround each chamber. of NH3N loss trapped by the chambers. A function was fit
Water was applied daily to the plates during field calibration to using a linear regression procedure in the graphics program
maintain the soil near field capacity, but dry enough to avoid N (SigmaPlot version 11.0, SPSS Inc., Chicago, IL).
loss by denitrification. A sulfuric acid solution (50 mL) [1 mol L1 In 2010, the air temperature and relative humidity,
H2SO4 solution plus 4.0% (v/v) glycerol (C3H8O)] was used in precipitation, and wind conditions were recorded by an automated
the foam NH3N traps inside the chambers. Before being soaked weather station located within 200 m of the plot area by the
in the acid trap solutions, the foam discs and strips were washed Colorado Agricultural Meteorological Network (CoAgMet,
twice with distilled water, then wrung-out and squeezed to remove 2011). The wind run (the distance of air flow past a point during
as much water as possible. We assumed that NH3 was the major a day) at 203 cm height aboveground level was reported to show
form of N loss during chamber calibration. the average wind intensity during each day and average wind gust
The total amount of NH3N loss from each of three N speeds recorded each day at the weather station.
rates was determined by a 15N balance method. The N rates
were 50, 100, and 200 kg N ha1, designated N1, N2, and N3, Nitrogen Source Studies (2010 and 2011)
respectively, using 15N labeled urea with 1.55% atom excess Ammonia emissions from two commonly used N fertilizer
with four replications. The treatments consisted of plates of soil sources, dry granular urea (46%N) and liquid urea ammonium
with or without chambers present above the soil to evaluate the nitrate (UAN 32%N), were compared to two controlled
influence of chambers on NH3N loss. The control treatment release N fertilizers (SuperU [46%N] and ESN [44%N]).
(without a chamber) was used to evaluate the absolute amount The controlled-release, polymer-coated urea (ESN), gradually
of NH3N loss. The chamber calibration sampling occurred 4, releases N during the growing season. The stabilized urea source,
7, and 11 d after N fertilization. SuperU (registered product of KOCH Agronomic Services,
Following the last sampling, the soil surface was acidified Witchita, KS), contains urease [N-(n-butyl)-thiophosphoric
to terminate NH3N volatilization losses with 80 mL of 0.5 triamide] (NBPT) and nitrification (dicyandiamide) inhibitors
M of H2SO4. The soil from each treatment was air dried (72 that are uniformly distributed throughout the granule during
h), ground, and mixed. Soil samples were sieved through a 2 the manufacturing process (Halvorson et al., 2010). This
mm screen before preparation for soil analyses. Soil samples experiment evaluated the effectiveness of both chamber types
were ground to pass a 150-m screen and duplicate samples in recovering ammonia emissions from these N sources in
of each treatment analyzed for total soil N and 15N content small plot areas under field conditions, and used the calibration
using a Carlo-Erba automated C/N analyzer (Haake Buchler factors determined in the chamber calibration experiment to
Instruments, Inc., Saddle Brook, NJ) interfaced to a tracer estimate total NH3N loss from each N source. The N sources
Table 1. Amount of 15N labeled urea solution applied to soil on each plate for each N rate treatment, soil 15N excess concentration,
total soil N derived from fertilizer, and total and percent N loss from fertilizer for each chamber type and control during field cali-
bration of each chamber type.
Chamber N Rate N applied Soil 15N excess Soil N derived from fertilizer N loss from fertilizer N loss
mg N plate1 % mg N plate1 %
Semi-open N1 117.6 0.0971e 64.4d 53.2d 45.2bc
Semi-open N2 235.0 0.1546d 109.9c 125.1b 53.2a
Semi-open N3 470.1 0.3063b 258.6a 211.5a 45.0bc
Open N1 117.6 0.0961e 63.7d 53.8d 45.8bc
Open N2 235.0 0.1673cd 119.1Bc 115.9bc 49.3ab
Open N3 470.1 0.3094ab 261.2a 209.7a 44.5bc
Control (no chamber) N1 117.6 0.1034e 67.4d 50.1d 42.6c
Control (no chamber) N2 235.0 0.1814c 126.9b 108.1c 46.0bc
Control (no chamber) N3 470.1 0.3252a 270.2a 199.9a 42.5c
CV, % 6.05 6.81 8.20 8.15
N rates applied: N1 = 50 kg ha 1; N2 = 100 kg ha 1; N3 = 200 kg ha 1.
Values within a column followed by the same letter are not significantly different at = 0.05 (LSD).

1598 Agronomy Journal Volume 104, Issue 6 2012

were manually band-applied (5 cm width) on the soil surface (Analytical Software, Tallahassee, FL) to determine treatment
about 10 cm from corn row shortly after corn emergence at a effects for each year. When comparing N sources across years,
rate of 200 kg N ha1. Immediately following irrigation, all the ANOVA was conducted across chambers and N sources with
chambers were fitted with acid traps and placed on the soil years included within replications. All statistical comparisons were
surface over the fertilizer band within each treatment. Rubbers made at the 0.05 probability level unless otherwise stated using the
bands were placed around the chamber and stakes to support least significant difference (LSD) method for mean separation.
the chambers in an upright position.
In 2010, the N source study was conducted in the same Results and Discussion
check plot area used for the chamber calibration. Both Calibration of Ammonia Chambers
chamber types were placed adjacent to each other near the The chamber calibration work was initiated a few days
plot center in the urea and SuperU treatments in 2010. Check after the initiation of the N source study in 2010 to avoid the
treatments (no N applied) were located in the corn row application of irrigation water to the soil on the calibration plates
between N sources plots. In 2010, both chamber types were and possible loss of 15N labeled soil. During the 15N balance field
sampled six times (Day 3, 7, 10, 14, 21, and 27 with additional experiment (1728 June 2010) to calibrate chamber efficiencies,
sampling days for the check and ESN treatments, Days 37, 44, the maximum and minimum temperature were 28.4 and 11.5C,
50, and 57) after N application on 25 May. On 26 May 2010, respectively, relative humidity was 22.4%, with wind gusts
16 mm of irrigation water was applied to the plot area with a reaching an average of 29.7 km h2 , and the average wind run
linear move, sprinkler irrigation system. The extended days was 200 km d1 (Fig. 3). These weather observations were similar
of sampling for the check (no N applied) and ESN were to to the past 5 yr seasonal average during May and June as recorded
investigate if there would be any NH3 volatilization from ESN by the weather station.
that would correspond with increased nitrous oxide emissions Both chamber types showed similar recovery patterns for
observed with ESN in previous studies at about 40 d after NH3N loss between sampling dates and cumulative NH3N
ESN application (Halvorson et al., 2010, 2011). Both chamber loss amounts with time (Fig. 4). For both chamber types, the
types were used to measure NH3N loss from the check (no highest NH3N losses between sampling dates were observed
N fertilizer applied), urea, and SuperU treatments. The UAN
and ESN treatments were also evaluated using the semi-open
chambers to measure NH3N loss. Within each replication
(four in 2010), the plot size for each N source was one corn
row 1 m in length (76 cm row spacing) with 0.4 m alleyways
between replications. The chambers were secured in the
upright position using wire stakes as described for the chamber
calibration study, and installed 2 d before N fertilization.
In 2011, the N source study was conducted in plots that
had received 200 kg N ha1 for the previous two cropping
seasons. Only the open-chamber type was used in 2011, and
they were located in the band applied N plots used in an N
source placement study on nitrous oxide emissions with three
replications (A.D. Halvorson and S.J. Del Grosso, unpublished
data, 2012). The plot replicates were spread out over a much
larger area in 2011 than in 2010. In 2011, the open-chambers
were sampled 13 times (Day 6, 9, 12, 16, 22, 29, 34, 37, 42, 47,
54, 61, and 68) after N application on 25 May 2011. On 26
May 2011, 19 mm of irrigation water was applied with a linear
move, sprinkler irrigation system to the entire plot area.
Ammonia volatilization emission by N sources was expressed
as the NH3N loss between sampling dates (g NH3N ha1)
and the cumulative amount over the entire sampling period (g
NH3N ha1). The NH3N loss between sampling dates was
calculated on an area basis as a weighted average of the NH3N
emission from the area under the chamber covering the fertilizer
band and the NH3N emission from the check area without
fertilizer between corn rows. The NH3N loss between sampling
dates for each chamber was determined using the calibration
factors determined in the 15N chamber calibration experiment.
All treatments were arranged as a randomized complete block
design with four replications in 2010 and three replications
in 2011. Plot size in 2011 was 3 by 4.6 m for each N source
within each replication. Randomized complete block ANOVA Fig. 3. Air temperature, relative humidity, precipitation, wind
were performed using Analytical Software Statistix9 program gust, and wind run during field chamber calibration study in 2010.

Agronomy Journal Volume 104, Issue 6 2012 1599

4 d after N fertilization for the N1 and N2 treatments and 7 d
after fertilization for the N3 treatment.
The 15N excess results showed a significant difference for
N2 under semi-open chamber compared with the N2 control
treatment with no chamber present (Table 1). Total N derived
from fertilizer remaining in the soil was different for N2
treatment under semi-open chambers. The absolute amounts
of NH3N loss observed in control treatments (no chamber
present) were 50.1 mg NH3N plate1 from N1, 108.1 mg
NH3N plate1 from N2, and 199.9 mg NH3N plate1 from
N3 which represented around 42.6, 46, and 42.5% of the total N
applied, respectively. There were no significant differences in the
absolute amounts of NH3N loss and total NH3N loss with
open chambers at all N fertilizer rates, and for the N1 and N3
rates with the semi-open chambers. The NH3N losses under
the control treatments (no chamber present) and open-chamber
treatments were similar to those reported by Arajo et al. (2009).
The semi-open chamber recovered around 90, 63, and 65% of
NH3N loss for the N1, N2, and N3 treatments, respectively
(Table 1, Fig. 4). The open-chambers recovered 100, 76,
and 79% of NH3N losses for N1, N2, and N3 treatments,
respectively (Table 1, Fig. 4). These data would suggest a
slightly greater recovery of NH3 loss at the lowest N rate than
for the N2 and N3 N rates.
Fig. 4. Cumulative NH3 N loss from (A) 15N labeled urea with
time and (B) NH3 N loss between sampling dates as trapped
by the semi-open and open chambers during calibration in
2010. Vertical lines in B represent standard error values.
(N1 = 50 kg N ha 1; N2 = 100 kg N ha 1; N3 = 200 kg N ha 1).
1600 Agronomy Journal Volume 104, Issue 6 2012
The difference between maximum and minimum air
temperature was 15 to 20C (Fig. 3). The relative humidity
reached around 50 to 60% some days and dew occurred naturally,
maintaining soil moisture. Wind flow and intensity were
important factors. Without the use of soil anchors, there appeared
to be more air flow around and exchange inside the chambers,
helping to promote NH3N capture with this air movement.
The correlation between NH3N loss determined by 15N
balance and NH3N trapped by the chambers in the field
resulted in similar regression fits and calibration factors for
both chamber types (Fig. 5). Development of calibration
factors for each chamber type under field condition resulted in
consistent NH3N loss measurements for both chamber types.
Therefore, we chose to use only the open-chamber type in the
2011 N source study. The open-chamber was easier to use than
the semi-open chamber.
Nitrogen Source Studies
The check and urea treatments showed similar patterns of
NH3N losses between sampling dates and cumulative NH3N
losses with time for both chamber types (Fig. 6) in 2010. For the
urea treatment, the highest fluxes occurred 7 d after fertilization,
similar to the patterns observed for the same N rate (N3
treatment, 200 kg N ha1) in the 15N calibration experiment.
The exception was NH3N fluxes and cumulative losses
under SuperU. With the semi-open chambers, we observed an
increased NH3N flux at 14 and 21 d after N fertilization. With
the open-chambers, we observed a smaller increase in NH3N
flux for these same days, but fluxes were more similar to that of
the check treatment (no N added). The authors do not have an
explanation for the differences in NH3N fluxes from SuperU
between the two chamber types. Some studies have shown that
NH3N loss from NBPT treated N fertilizers (urease inhibitor)
can depend on soil type and air humidity conditions (Rawluk
et al., 2001; Cantarella et al., 2008). It is possible that the semi-
open chamber had a higher humidity level in the area between
the soil and the lower foam (Fig. 1) in the semi-open chamber
than was present in the open-chamber design. Humidity
Fig. 5. Correlation between absolute amounts of NH3 N loss
as determined by 15N applied in control treatment (without
chamber influence) and total NH3 N loss trapped by chambers
(semi-open and open) in 2010 chamber calibration experiment.
measurements were not made to confirm that this might have
happened. The second year (2011) of measurements with the
open-chamber showed low NH3N fluxes for SuperU (Fig. 8a),
similar to those observed in 2010 for the open-chamber (Fig. 6a).
Using the semi-open chamber, NH3N loss between sampling
dates and cumulative NH3N loss with time varied with N source
(Fig. 7). The highest NH3N fluxes were observed from UAN,
at 7 d after N fertilization. With ESN, the largest fluxes occurred
42 d after fertilization, as expected based on N2O emissions
patterns (Halvorson et al., 2011), because the polymer-coated urea
promotes a later N release from the fertilizer granule.
In 2011, the NH3N loss between sampling dates was
highest with UAN on sampling Day 6, with UAN and
urea being significantly higher than the other N sources on
sampling Days 12 and 22; and with ESN being significantly
higher than the other N treatments on sampling Day 47, 61,
and 68 (Fig. 8b). The cumulative NH3N losses with time
for each N source and check treatments are shown in Fig. 8a.
The cumulative NH3N loss pattern for each N treatment
measured with the open-chamber in 2011 was similar to the
cumulative NH3N loss pattern with time observed with the
semi-open chamber in 2010 (Fig. 7a), with the exception of
SuperU. However, the NH3N loss pattern with time in 2010
observed with the open-chamber for SuperU (Fig. 6a) was
similar to that in 2011. As was the case in 2010, we observed
increased NH3N emissions with ESN starting with the
Fig. 6. (A) Cumulative NH3 N loss with time and (B) NH3 N
loss between sampling dates from check (no N applied), urea,
and SuperU recovered by semi-open and open chambers in
2010. Vertical lines in B represent standard error value.
Agronomy Journal Volume 104, Issue 6 2012 1601
Day 42 sampling through the end of the 2011 measurement
period (Day 68). The higher NH3N emissions in 2011 than
in 2010 possibly occurred because the plots in 2010 were
located on a check treatment that had not had any N fertilizer
applied for 11 yr. The plot area used in 2011 had >134kg N
ha1 applied each year during the same 11 yr period, with 200
kg N ha1 applied during the previous 2 yr. The check (no N
applied) in 2011 also had a much higher NH3N emission
level than the check in 2010. The NH3N loss above that of
the check treatment in 2011 is reported in Table 2. Variability
between replications was high for ESN during the last four
sampling dates in 2011 (Fig. 8b), thus the differences shown
are marginally significant at = 0.2 (P = 0.17), with SuperU
having a lower NH3N loss than UAN and ESN, which is
consistent with the 2010 data. Excluding the ESN data from
the ANOVA analysis resulted in a significant difference
among N sources at an = 0.06 (P = 0.056) with UAN being
significantly greater than SuperU and urea not being different
from UAN or SuperU. The mean separation among UAN,
Urea, and SuperU was the same as that shown in Table 2
when including ESN in the analysis. Some of the variability in
NH3N loss during the last four sampling periods with ESN
in 2011 was possibly caused by hot sun rays causing the foam
strips (Fig. 2) to shrivel slightly in some chambers (random
effect) which possibly decreased the capacity of the acid foam
strip to effectively absorb NH3. This problem was eliminated in
Fig. 7. Semi-open chamber measurement of (A) cumulative
NH3 N loss with time and (B) NH3 N loss between sampling
dates for check (no N applied), urea, ureaammonium nitrate
(UAN), ESN, and SuperU treatments in 2010. Vertical lines in
B represent standard error value.
a subsequent study by coating the outside of the open-chamber
with silver paint to reflect the suns rays.
Combining and analyzing data from both years revealed
that there was no significant year effect on NH3N loss or a
year N source interaction. Differences among N sources were
significant at P < 0.001 when analyzed across years, with ESN
and UAN having a greater NH3N loss than urea, and urea a
greater loss than SuperU (Table 2).
Ammonia loss from each N source over that of the check
treatment (no N added) ranged between 0.1 and 3.6% of total
N applied for both chamber types in 2010 and 0.3 to 4.0% in
2011 (Table 2). The urea NH3N losses observed in this study
(1.9 and 2.3% observed in open and semi-open chambers,
respectively, in 2010 and 2.4% in 2011) were low compared to
the 7% average loss from urea (review of 148 studies) reported
by Bouwman et al. (2002) for temperate climate countries. The
UAN NH3N loss was 3.3 to 3.7% and was within the range of
0.3 to 13% reported by Ma et al. (2010), when 150 kg N ha1 was
applied. In our study, the irrigation water applied the day after
N fertilization probably reduced NH3 volatilization loss from
urea and UAN. Holcomb et al. (2011) reported that application
of at least 13 mm of irrigation water through a sprinkler
system could reduce NH3 losses from urea to <3%. Measured
with the semi-open chamber, SuperU reduced NH3N losses
32% compared to urea in 2010, and measured with the open-
chambers, 95% in 2010 and 88% in 2011. Ammonia-N losses
from ESN and UAN were 38 and 37%, respectively, greater
than from urea in 2010 and 67 and 54% in 2011, respectively.
Urea NH3N losses measured with semi-open chambers were
not statistically different from the other N sources in 2010;
however, NH3N losses from urea were significantly greater
than from SuperU with the open chambers in 2010 and followed
a similar trend in 2011. Rochette et al. (2009) observed that
NH3N losses were 9, 4, and 5% of total N applied, from urea,
ESN, and urea treated with NBPT, respectively, in broadcast
N fertilization applications. Pereira et al. (2009) showed that
urease inhibitor in either polymer-coated urea or urea reduced
NH3N volatilization by 50% compared to urea in a side-dress
N fertilization study, but their measurements were only for 11 d
after N fertilization. Our results suggest that it may be important
to extend the measurement period when evaluating NH3N
losses from polymer-coated urea fertilizers to monitor NH3
losses when the polymer-coated granule starts releasing urea
N to the crop. It is possible that the NH3 emissions from ESN
granules which remained on the soil surface would have been
lower had the chambers been removed during irrigation. The
irrigation water may have moved the urea that had defused from
the polymer-coated granule during irrigation into the soil, which
may reduce NH3 loss.
Consistent with the study by Holcomb et al. (2011),
application of irrigation water the day after N fertilization
suppressed the loss of NH3N from these urea based fertilizers
to <4% of N applied. If irrigation water is readily available,
one might consider using irrigation as a management tool to
control NH3N loss rather than controlled release fertilizers
which are more expensive than urea and UAN. In our studies
on nitrous oxide emissions from different N sources, we have
not observed a corn grain yield advantage with the controlled
release N sources over that of urea or UAN, thus limiting
any economic incentive to use the more expensive N sources,

Table 2. Cumulative NH3 N loss for each fertilizer source

above that of the check (no N applied) and loss per kg N ap-
plied (200 kg N kg N ha 1) for the 2010 and 2011 measurement
periods and across both years.
N NH3N loss Loss of
Chamber source from fertilizer N applied
kg N ha1 %
2010
Semi-open Urea 5.2ab 2.6
Semi-open SuperU 3.7b 1.9
Semi-open ESN 7.2a 3.6
Semi-open UAN 7.2a 3.6
Open Urea 3.8b 1.9
Open SuperU 0.1 c 0.1
2011
Open Urea 4.8ab 2.4
Open SuperU 0.6b 0.3
Open ESN 8.0a 4.0
Open UAN 7.4a 3.7
Combined 2010 and 2011 Data
Urea 4.6b 2.3
SuperU 1.6c 0.8
ESN 7.6a 3.8
Fig. 8. Open chamber measurement of (A) cumulative NH3 N UAN 7.3a 3.7
loss with time and (B) NH3 N loss between sampling dates for Values within this column for 2010 and combined years followed by the same
check (no N applied), urea, ureaammonium nitrate (UAN), letter are not significantly different at = 0.05 (LSD).
ESN, and SuperU treatments in 2011. Vertical lines in B Values within this column for 2011 followed by the same letter are not signifi-
represent standard error value. cantly different at = 0.2 (LSD).

1602 Agronomy Journal Volume 104, Issue 6 2012


such as SuperU which contains a urease inhibitor, to reduce
NH3N losses. However, the controlled release N sources have
been effective in our semiarid environment in reducing nitrous
oxide emissions up to 70% when compared to urea (Halvorson
et al., 2010, 2011). Thus, the controlled release N sources can be
a management option for reducing nitrous oxide emissions and
protecting the environment.
CONCLUSIONS
The 15N balance technique method was useful in establishing
reliable calibration factors for both chamber types, under field
conditions. Both semi-static chambers (semi-open and open)
showed similar NH3N loss results during the calibration study
and the N source study in 2010. Immediate irrigation following
N fertilization appears to have reduced NH3N volatilization
losses (range 2.33.7% of N applied) from traditional N
fertilizers (urea and UAN). Lowest NH3N losses (0.8% of N
applied) were observed with SuperU in this 2 yr study. Highest
NH3N losses (3.73.8%% of N applied) were observed with
UAN and ESN. The results indicate that NH3N measurement
times (days) may need to be extended to evaluate NH3N
volatilization losses from controlled release N source, such as
polymer-coated urea. The open-chamber method reported by
Arajo et al. (2009) appears to be an inexpensive and reasonably
accurate method to assess NH3N losses from N fertilization
studies in a small plot experiments.
ACKNOWLEDGMENTS
The authors thank E. Buenger, P. Norris, B. Floyd, and R. DAdamo
for their assistance and support in collecting, processing, and analyzing
the ammonia solutions, soil samples, and data reported herein, and the
ARDEC staff for help with plot maintenance and Dr. Mark West, ARS
Statistician, for help with the statistical analysis. Dr. Jantalia thanks
EMBRAPA for the financial support and USDA-ARS for providing
the facilities for all the experimental steps. The authors thank the
IPNI Foundation for Agronomic Research with funding from Agrium
Inc. and Agrotain International for financial support for this project.
This publication is also based on work supported by the Agricultural
Research Service under the ARS GRACEnet Project.
REFERENCES
Aneja, V.P., J. Blunden, K. James, W.H. Schlesinger, R. Knighton, W. Gilliam et al. 2008.
Ammonia assessment from agriculture: U.S. status and needs. J. Environ. Qual. 37:515
520 doi:10.2134/jeq2007.0002in
Aneja, V.P., D. Nelson, P. Roelle, and J. Walker. 2003. Agricultural ammonia emissions and
ammonium concentrations associated with aerosols and precipitation in the southeast
United States. J. Geophys. Res. 108 (D4), 4152. p. 12-1 to 12-11. doi:10.1029/2002JD002271
Arajo, E.S., T. Marsola, M. Miyazawa, L.H.B. Soares, S. Urquiaga, R.M. Boddey, and B.J.R.
Alves. 2009. Calibration of a semi-opened static chamber for the quantification of
volatilized ammonia from soil. (In Portuguese, with English abstract.) Pesqi. Agropecu.
Bras. 44:769776 doi:10.1590/S0100-204X2009000700018
Bouwman, A.F., L.J.M. Boumans, and N.H. Batjes. 2002. Estimation of global NH3 volatilization
loss from synthetic fertilizers and animal manure applied to arable lands and grasslands.
Global Biogeochem. Cycl. 16:81, 813. doi:10.1029/2000GB001389
Bouwmeester, R.J.B., P.L.G. Vlek, and J.M. Stumpe. 1985. Effect of environmental factors on
ammonia volatilization from a urea-fertilized soil. Soil Sci. Soc. Am. J. 49:376381.
doi:10.2136/sssaj1985.03615995004900020021x
Cabezas, A.R.L., P.C.O. Trivelin, J.A. Bendassolli, D.G. Santana, and G.J. Gascho.
1999. Calibration of a semi-open static collector for determination of ammonia
volatilization from nitrogen fertilizers. Commun. Soil Sci. Plant Anal. 30:389406.
doi:10.1080/00103629909370211
Cabrera, M.L., D.E. Kissel, R.C. Davis, N.P. Qafoku, and W.I. Segars. 2001. Design and
ammonia-recovery evaluation of a wind speed-sensitive chamber system. Soil Sci. Soc. Am.
J. 65:13021306. doi:10.2136/sssaj2001.6541302x
Cantarella, H., P.C.O. Trivelin, T.L.M. Contin, F.L.F. Dias, R. Rossetto, R. Marcelino, R.B.
Coimbra, and J.A. Quaggio. 2008. Ammonia volatilisation from urease inhibitor treated
Agronomy Journal Volume 104, Issue 6 2012 1603
urea applied to sugarcane trash blankets. Sci. Agric. (Piracicaba, Brazil) 65:397401.
doi:10.1590/S0103-90162008000400011
Clay, D.E., G.L. Malzer, and J.L. Anderson. 1990. Ammonia volatilization from urea as influenced
by soil temperature, soil water content, and nitrification and hydrolysis inhibitors. Soil Sci.
Soc. Am. J. 54:263266. doi:10.2136/sssaj1990.03615995005400010042x
CoAgMet. 2011. Colorado Agricultural Meteorological net work. Colorado State Univ. www.
coagmet.com (accessed 1 June 2011).
FAO. 2006. Plant nutrition for food security: A guide for integrated nutrient management.
Fertilizer Plant Nutr. Bull. 16. Food and Agriculture Organization of the United Nations.
ftp://ftp.fao.org/agl/agll/docs/fpnb16.pdf (accessed 28 Aug. 2012).
Ferguson, R.B., and D.E. Kissel. 1986. Effects of soil drying on ammonia volatilization from
surface-applied urea. Soil Sci. Soc. Am. J. 50:485490 10.2136/sssaj1986.0361599500500
0020047x. doi:10.2136/sssaj1986.03615995005000020047x
Galloway, J.N., F.J. Dentener, D.G. Capone, E.W. Boyer, R.W. Howarth, S.P. Seitzinger et al.
2004. Nitrogen cycles: Past, present, and future. Biogeochem. 70:153226. doi:10-1007/
s10533-0370-0.
Grant, C.G., S. Jia, K.R. Brown, and L.D. Bailey. 1996. Volatile losses of NH3 from surface-
applied urea and urea ammonium nitrate with and without the urease inhibitors NBPT or
ammonium thiosulphate. Can. J. Soil Sci. 76:417419. doi:10.4141/cjss96-050
Halvorson, A.D., S.J. Del Grosso, and F. Alluvione. 2010. Tillage and inorganic nitrogen source
effects on nitrous oxide emissions from irrigated cropping systems. Soil Sci. Soc. Am. J.
74:436445. doi:10.2136/sssaj2009.0072
Halvorson, A.D., S.J. Del Grosso, and C.P. Jantalia. 2011. Nitrogen source effects on soil nitrous
oxide emissions from strip-till corn. J. Environ. Qual. 40:17751786. doi:10.2134/
jeq2011.0194
Holcomb, J.C., D.M. Sullivan, D.A. Horneck, and G.H. Clough. 2011. Effect of irrigation rate on
ammonia volatilization. Soil Sci. Soc. Am. J. 75:23412347. doi:10.2136/sssaj2010.0446
Janzen, H.H., and S.M. McGinn. 1991. Volatile loss of nitrogen during decomposition of legume
green manure. Soil Biol. Biochem. 23:291297. doi:10.1016/0038-0717(91)90066-S
Lara-Cabezas, W.A.R., and P.C.O. Trivelin. 1990. Eficincia de um coletor semi-aberto esttico
na quantificao de N-NH3 volatilizado da uria aplicada ao solo. (In Portuguese, with
English abstract.) Rev. Bras. Cienc. Solo 14:345352.
Ma, B.L., T.Y. Wu, N. Tremblay, W. Deen, N.B. McLaughlin, M.J. Morrison, and G. Stewart.
2010. On-farm assessment of the amount and timing of nitrogen fertilizer on ammonia
volatilization. Agron. J. 102:134144. doi:10.2134/agronj2009.0021
McGinn, S.M., and H.H. Janzen. 1998. Ammonia sources in agriculture and their measurement.
Can. J. Soil Sci. 78:139148. doi:10.4141/S96-059
Misselbrook, T.H., and M.N. Hansen. 2001. Field evaluation of the equilibrium concentration
technique (JTI method) for measuring ammonia emission from land spread manure or
fertilizer. Atmos. Environ. 35:37613768. doi:10.1016/S1352-2310(01)00169-8
Misselbrook, T.H., F.A. Nicholson, B.J. Chambers, and R.A. Johnson. 2005. Measuring ammonia
emissions from land applied manure: An intercomparison of commonly used samplers and
techniques. Environ. Pollut. 135:389397. doi:10.1016/j.envpol.2004.11.012
Misselbrook, T.H., T.J. Van Der Weerden, B.F. Pain, S.C. Jarvis, B.J. Chambers, K.A. Smith et
al. 2000. Ammonia emission factors for UK agriculture. Atmos. Environ. 34:871880.
doi:10.1016/S1352-2310(99)00350-7
Nmmik, H. 1973. The effect of pellet size on the ammonia loss from urea applied to forest soils.
Plant Soil 39:309318. doi:10.1007/BF00014798
Pacholski, A., G. Cai, R. Nieder, J. Richter, X. Fan, Z. Zhu, and M. Roelcke. 2006. Calibration
of a simple method for determining ammonia volatilization in the fieldComparative
measurements in Henan Province, China. Nutr. Cycl. Agroecosyst. 74:259273.
doi:10.1007/s10705-006-9003-4
Pereira, H.S., A.F. Leo, A. Verginassi, and M.A.C. Carneiro. 2009. Ammonia volatilization
of urea in the out-of-season corn. Rev. Bras. Cienc. Solo 33:16851694. doi:10.1590/
S0100-06832009000600017
Rawluk, C.D.L., C.A. Grant, and G.J. Racz. 2001. Ammonia volatilization from soils fertilized
with urea and varying rates of urease inhibitor NBPT. Can. J. Soil Sci. 81:239246.
doi:10.4141/S00-052
Rochette, P., D.A. Angers, M.H. Chantigny, J.D. MacDonald, M. Gasser, and N. Bertrand. 2009.
Reducing ammonia volatilization in a no-till soil by incorporating urea and pig slurry in
shallow bands. Nutr. Cycl. Agroecosyst. 84:7180. doi:10.1007/s10705-008-9227-6
Rochette, P., D. Guilmette, M.H. Chantigny, D.A. Angers, J.D. MacDonald, N. Bertrand et al.
2008. Ammonia volatilization following application of pig slurry increases with slurry
interception by grass foliage. Can. J. Soil Sci. 88:585593. doi:10.4141/CJSS07083
Sherlock, R.R., J.R. Freney, N.P. Smith, and K.C. Cameron. 1989. Evaluation of a sampler
for assessing ammonia losses from fertilized fields. Fert. Res. 21:6166. doi:10.1007.
BF01080531
Snyder, C.S. 2008. Fertilizer nitrogen BMPs to limit losses that contribute to global warming. Int.
Plant Nutrition Inst., Norcross, GA.
Sommer, S.G., and B.T. Christensen. 1992. Ammonia volatilization after injection of anhydrous
ammonia into arable soils of different moisture levels. Plant Soil 142:143146. 10.1007/
BF00010184
Stewart, W.M. 2008. Fertilizer sources for irrigated corn. In: W.M Stewart and W.B. Gordon,
editors, Fertilizing for irrigated corn: Guide to best management practices. IPNI no.
30-3240. Int. Plant Nutrition Inst., Norcross, GA. p. 2-1 to 2-6.
Zhang, Y., A.J. Dore, L. Ma, X.J. Liu, W.Q. Ma, J.N. Cape, and F.S. Zhang. 2010. Agricultural
ammonia emissions inventory and spatial distribution in the North China Plain. Environ.
Pollut. 158:490501. doi:10.1016/j.envpol.2009.08.033