Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Contents lists available at ScienceDirect
Agricultural Water Management
journal homepage: www.elsevier.com/locate/agwat
Impacts of nitrogen application timing and cover crop inclusion on
subsurface drainage water quality
Michael D. Ruffatti
a
, Richard T. Roth
b
, Corey G. Lacey
b
, Shalamar D. Armstrong
b,⁎
a
Department of Agriculture, Illinois State University, 301 North Main Street, Normal, IL 61790-5020, USA
b
Department of Agronomy, Purdue University, 915 West State Street, West Lafayette, IN 47907-2053, USA
ARTICLE INFO
Keywords:
In-field conservation practices
Nutrient loss reductions
Cereal rye
Daikon radish
Nitrate leaching
Unfertilized control
ABSTRACT
Significant reductions in nitrogen loading from sub-surface drainage fields of the Upper Mississippi River Basin
to the Gulf of Mexico will most likely be achieved from the mass adoption of nutrient loss reduction strategies at
a watershed scale. Few studies have quantified the efficacy of cover crops to reduce NO
3
-N loading in nitrogen
fertilizer management systems, where the dominant portion of the N rate is applied in the spring or fall, both of
which are common practices in the Upper Mississippi River Basin. In this experiment we quantified the impact of
N application timing and cover crop inclusion on NO
3
-N loss (leaching) from agricultural sub-surface drainage
within five nitrogen management scenarios: a zero control, applying the dominant portion of the N rate in the
spring, applying the dominant portion of the N rate in the fall, augmenting the a spring and Fall N application
system with cover crop. Each of the five nitrogen management scenarios was replicated three times on in-
dividually monitored sub-surface drainage plots established in Lexington, IL. During the experiment, a cereal rye
(Secale cereal L.) and radish (Raphanus sativus L.) blend was interseeded within both corn (Zea mays L.) and
soybean (Glycine max L.). Fertilizer N application timing did not affect cover crop growth or N uptake. The
inclusion of cover crop resulted in more consistent and greater NO
3
-N loss reductions relative to adjusting
fertilizer N application timing from fall to spring. Cover crop reduced the flow-weighted NO
3
-N concentrations
by 39% and 38% and the N load by 40% and 47% when added to spring and fall fertilizer N management
systems, respectively. Cover crop proved to be effective in reducing NO
3
-N loss through sub-surface drainage
across the spectrum of N fertilizer management systems common to the Upper Mississippi River Basin.
1. Introduction
Inorganic fertilizer nitrogen (N) management for row crop produc-
tion only affects a minute percentage of the soils total N; however,
within exported tile drainage-water inorganic N is a significant pro-
portion of the total N (Blesh and Drinkwater, 2014). Furthermore, low
fertilizer N efficiency of row crops combined with high tile drainage
density in the Upper Mississippi River Basin (UMRB) contribute to the
export of excessive N, local water quality issues and the hypoxic zone in
the Gulf of Mexico (Gardner and Drinkwater, 2009;Smil, 1999). The
severity of this N loading issue resulted in the United States Environ-
mental Protection Agency Gulf of Mexico Hypoxia Task Force requiring
UMRB states to develop a Nutrient Loss Reduction Strategy (NLRS) to
reduce N and P loading (Mississippi River/Gulf of Mexico Watershed
Nutrient Task Force, 2008). The NLRS Science Assessment of each
UMRB state estimated that the manipulation of N application timing
and rate result in an N loading reduction of 10–20% on tile-drained
land, while cover cropping alone was estimated to reduce N loading by
28–40% (David et al., 2013;Illinois Nutrient Loss Reduction Strategy,
2015;Iowa Nutrient Reduction Strategy, 2013;Minnesota Nutrient
Reduction Strategy, 2014;Missouri Nutrient Loss Reduction Strategy,
2014;Wisconsin Nutrient Reduction Strategy, 2013). Among all UMRB
NLRS, cover crops demonstrated the highest efficacy to achieve the
proposed non-point source nutrient loss reduction targets on a wa-
tershed scale.
The scientific literature has demonstrated that actively growing
cover crops (CC) influence the concentration of nitrate (NO
3
-N) in tile
water during the fallow period of the year, which frequently results in
less nitrate loading (Drury et al., 2014;Kaspar et al., 2007,2012;Strock
et al., 2004). CC absorb inorganic N from the residual, fertilized and
mineralized soil N pools affecting the distribution of inorganic N in the
soil profile (Kaspar et al., 2007;Lacey and Armstrong, 2014). The
presence of both winter-kill and winter hardy CC species result in sig-
nificantly less soil inorganic N at lower soil depths closer to the location
https://doi.org/10.1016/j.agwat.2018.09.016
Received 24 January 2018; Received in revised form 31 August 2018; Accepted 6 September 2018
⁎
Corresponding author.
E-mail address: sarmstro@purdue.edu (S.D. Armstrong).
Agricultural Water Management 211 (2019) 81–88
Available online 02 October 2018
0378-3774/ © 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
T
of the tile drainage (Cooper et al., 2017;Lacey and Armstrong, 2014).
Studies have also demonstrated that evapotranspiration from CC re-
duces the soil moisture content in the fallow period without negatively
affecting cash crop yield (Basche et al., 2016). This reduction in soil
moisture content has resulted in a higher soil matric potential, a lower
soil leaching potential, and less tile drainage resulting in reduced N
loading (Daigh et al., 2014;Kaspar et al., 2007;Qi and Helmers, 2010).
Few studies have quantified the environmental impacts of sys-
tematic conservation, where multiple nutrient loss reduction strategies
are concurrently applied to one field, such as N application timing and
CC. In spring N application systems, the ability of cover crops to im-
prove water quality has been documented in the literature. For ex-
ample, in Iowa, Kaspar et al. (2007) studied the system of spring N
management combined with CC inclusion and determined that a rye CC
significantly reduced the average annual flow-weighted NO
3
-N con-
centration of drainage water by 50% or more compared to the control.
In Minnesota, Strock et al. (2004) also studied the impact of estab-
lishing a cereal rye CC following corn, within a spring N application
system, and found a 13% reduction in NO
3
-N loading via tile drainage.
While fertilizer N applications have been trending toward the spring,
there remains a large percentage (41–46%) of row crop acreage in the
UMRB that receives fall-applied N (Bierman et al., 2012;Illinois
Nutrient Loss Survey Results, 2016;Lemke et al., 2011;Ribaudo et al.,
2012;Smiciklas et al., 2008). The source of N during this fall N appli-
cation could be anhydrous ammonia, ammonium phosphate or live-
stock manure. Equipment and labor availability in the fall, reduced N
fertilizer costs, and spring soil conditions have all been suggested as
reasons for fall N application (Ribaudo et al., 2012;Smiciklas et al.,
2008). To achieve the targeted surface area reduction of the Gulf of
Mexico Hypoxic Zone established by the USEPA, all agricultural ferti-
lizer N management systems common to the UMRB must significantly
improve N retention. Currently, there remains a dearth of research
concerning the ability of cover crops to reduce tile nitrogen con-
centrations within fall N application systems, and that investigates the
concurrent adoption of both N timing and CC. Therefore, the objectives
of this study were to quantify the impact of N application timing and CC
inclusion on the flow-weighted NO
3
-N concentration and loading from
agricultural tile drainage within five N management scenarios (i) ap-
plying the dominant portion of the N rate in the spring, (ii) applying the
dominant portion of the N rate in the fall, (iii) augmenting the a spring
N system with CC, (iv) and augementing the Fall N application system
with CC, and (v) a zero control without N fertilizer or CC.
2. Materials and methods
2.1. Site description and cultural practices
The experimental site was located east of Lexington, Illinois
(40°38′25.9″N 88°43′11.2″W) at the Illinois State University Nitrogen
Management Research Field Station. The predominant soil types within
the approximately 10 ha site are Drummer and El Paso (67.5%) and
Hartsburg (26%) silty clay loams, both soil types are common in the
central Illinois region and are classified as poorly drained with a 0–2%
slope. The drainage system was established on April 18, 2014. Three
7.6 cm inside diameter tile laterals spaced 13.7 m apart were installed
in each plot at an approximate depth of 0.9 m. The laterals merge 4.5 m
from a controlled drainage structure before connecting to 15.2 cm main
tile. Precipitation and air temperature data were collected from a
weather station located at the experimental site in each year of the
experiment.
The production history of this field consists of an eight-year rotation
of rain-fed strip-tilled corn (Zea mays L.) and no-till soybeans (Glycine
max L.), which were both harvested for grain. This experiment was a
continuation of these cultural practices. The site was comprised of fif-
teen individually tile-drained 0.65 ha plots, each of which included a
tile-water monitoring station. The experiment consisted of five
treatments replicated three times arranged in a complete randomized
block design. The experimental treatments included a zero control (no
N, no CC), a spring dominated nitrogen management system with and
without CC, and a fall dominated nitrogen management system with
and without CC (Table 1). All fall anhydrous ammonia (AA) was ap-
plied with a nitrification inhibitor, and application occurred only once
soil temperatures fell below 10 °C. The remaining N was applied as a
side-dress AA application, without an inhibitor, near the V6 growth
stage. Specific N sources and rates for each treatment can be found in
Table 1.
Corn and soybeans were planted in 76.2 cm rows using a John Deere
1770 N T 24-row planter. Corn was planted at a targeted rate of 86,485
seeds ha
−1
on April 30, 2015, and April 25, 2017. Population counts
resulted in average corn plant stands of 83,520 plants ha
−1
in 2015 and
87,990 plants ha
−1
in 2017. Soybeans were planted at a rate of 308,875
seeds per hectare on May 7, 2016. Weather conditions in the early
spring of the 2016 growing season caused poor emergence and resulted
in an average population of approximately 214,977 soybean plants per
hectare. Due to this reduction in the plant stand, a replant at a rate of
135,905 seeds per hectare occurred on May 25, 2016. After a popula-
tion check, the replant stand was found to be at approximately 133,434
plants per hectare, which resulted in an average of 348,411 plants per
ha. Harvest was conducted on September 23, 2015, October 21, 2016,
and October 9, 2017, using a John Deere S670 combine with a John
Deere 608C 8 row head for corn, and a John Deere 635FD 10.7-meter
flex draper head for soybeans (Deere & Company, Moline, Illinois, U.S.).
The CC mixture for this study was a 92% cereal rye (Secale cereal L.)
and 8% daikon radish (Raphanus sativus L.) blend calculated by weight,
first established in September 2014 and was grown in the same plots
each year. The CC were inter-seeded at a rate of 84 kg ha
−1
into the
standing crops using a Hagie STS12 (Hagie Manufacturing Company,
Clarion, Iowa, U.S.) modified with an air seeding box in early
September. Throughout the study, the daikon radish self-terminated
through vegetative desiccation in mid-to-late December following sev-
eral days of subfreezing weather conditions. The cereal rye, however, is
a winter hardy species that was chemically terminated at least two
weeks before the anticipated planting of the cash crop. Along with the
chemical termination of the CC, the research plots received pesticide
applications dependent upon the primary crop and weather conditions.
2.2. Cover crop shoot samples
CC sampling occurred in both the fall and spring to document both
above ground shoot biomass and nitrogen uptake. Within each treat-
ment, two 1 m
2
quadrants were randomly chosen, and the CC shoot
biomass was collected to create a representative sample for each
treatment. This sampling technique is a modified version of Dean and
Weil’s method developed in 2009 (Dean and Weil, 2009). In fall 2015
and fall 2016 radish and rye shoot biomass was separated and analyzed
by species. The CC biomass samples were oven dried at 60 °C and
ground to pass through a 1-mm sieve. The dry weight of each biomass
Table 1
Nitrogen Fertilizer source and nitrogen rate applied during the 2014–2015 and
2016–2017 corn years.
Fall Nitrogen System Spring Nitrogen System
2014–2015 2016–2017 2014–2015 2016–2017
Nitrogen Source kg N ha
−1
Fall Diammonium
Phosphate
40 50 40 50
Fall Anhydrous Ammonia 112 134 0 29
Spring Anhydrous
Ammonia
72 90 184 195
Total 224 274 224 274
M.D. Ruffatti et al. Agricultural Water Management 211 (2019) 81–88
82
sample was determined and used to calculate both total CC biomass, as
well as total CC nitrogen uptake. The dried and ground CC shoot bio-
mass were analyzed for percent total nitrogen using a 0.1000 g sample
via the use of a FLASH 2000 series dry combustion instrument (Ther-
moFisher Scientific, Waltham, Massachusetts, U.S.). The percent total
nitrogen was then multiplied by total CC biomass to determine total CC
nitrogen uptake (kg ha
−1
).
2.3. Corn and soybean yield
Grain yields were calculated by harvesting the entire plot area.
Grain weights were collected from a weigh wagon at harvest. A grain
subsample was then collected, wieghed and dried to measure grain
moisture content. Reported yields were corrected to 0.155 and 0.130 g
g
−1
moisture content for corn and soybean, respectively.
2.4. Water sampling
An automated tile water monitoring and sampling system was em-
ployed to determine the NO
3
-N flow-weighted concentration and
loading through the subsurface drainage system. The system included
an ISCO 6712 automated water sampling unit, an ISCO 2105 commu-
nication module, and an ISCO 2150 data logger module (Teledyne Isco,
Lincoln, Nebraska, U.S.), all of which were powered by a marine grade
12-volt battery maintained through the use of a solar panel and power
inverter. The automated sampler collected a 200 ml sample every hour
and formed a three-hour composite (600 ml) sample in each of the
twenty-four bottles. At the completion of the sampling program, each
plots hydrograph was analyzed and sampled to represent the base flow,
rising limb, peak flow, falling limb, and the inflection point of the hy-
drograph using the Flowlink software (Teledyne Isco, Lincoln,
Nebraska, U.S.). Each of the samples selected to be analyzed was fil-
tered with 0.45-micron filter paper to remove any suspended particu-
lates and analyzed for NO
3
-N concentrations using a Lachat QuikChem
®
8500 series flow injection analysis autosampler (Hach Company,
Loveland, Colorado, U.S.).
2.5. Statistical analysis
The experimental design was a complete random block with three
replications of five treatments. For each plot, the variables of average
flow-weighted NO
3
-N concentration, cumulative NO
3
-N load, cover
crop aboveground biomass, cover crop N uptake, and corn and soybean
yield were calculated individually by year. Data for all three years were
combined and the main effects of block, year, cover crop, and N ap-
plication timing were analyzed, along with the interactions of cover
crop by N application timing, cover crop by year, N application timing
by year, and cover crop by N application timing and year using the
PROC MIXED procedure (SAS 9.4, 2017). For all analyses, the
LSMEANS statement was used to apply the Tukey’s test at the 0.05
probability level to compare treatment means when the analysis of
variance indicated significant effects at the 0.05 alpha level.
3. Results
3.1. Weather data
The 2014 hydrologic year (the hydrologic year was determined
based on the average planting date of the cover crop, September-
August) was the coldest relative to 2015 and 2016, averaging 1.3 °C
cooler than the 30-year average. There was relatively little variation in
air temperature between the 2015 and 2016 hydrologic years, aver-
aging 1.2 and 1.1 °C above the 30-year average, respectively (Table 2).
To examine the influence of weather on different phases of the
cropping system, each hydrologic year was divided into cover crop
(September–April) and cash crop (May–August) growing seasons
(Fig. 1). Additionally, when examining the cover crop growing season,
both the fall and spring portions were considered separately. The
average ambient air temperatures for the fall portion (September–De-
cember) of the 2014 cover crop growing was slightly below the regional
norm, while the fall of 2015 and 2016 were slightly above the 30-year
average. Specifically, average air temperatures in November of 2015
and 2016 were on average 6.6 °C warmer than in 2014. In the spring
portion (January–April) of the cover crop growing seasons, an average
deviation from the 30-year norm of −2.1, +1.2, and + 2.9 °C was
observed in 2014, 2015, and 2016, respectively. In the spring of 2017,
January and February air temperatures were warmer than in spring
2015 and 2016. For example, the temperatures were 3.6 and 12.2 °C
warmer in January and February of 2017 compared to the same months
in 2015. Average monthly temperatures during the cash crop growing
season were similar to the 30-year norm and did not indicate any
limitation to corn or soybean growth.
Annual precipitation in 2014, 2015, and 2016 hydrologic years was
163.3 mm and 86.1 mm above the 30-year average and 221.0 mm
below average, respectively (Table 2). The fall cover crop growing
seasons had +137.5, +58.2, and −99.3 mm of precipitation, relative
to the 30-year average, in 2014, 2015, and 2016 respectively. Spring
cumulative precipitation deviated from the 30-year average by –127.1,
−86.7, and −32.1 mm in 2015, 2016, and 2017, respectively. Despite
having below average annual rainfall, the 2016 hydrologic year had the
greatest spring precipitation of all three years. Specifically, 180.6 mm
of cumulative rainfall in March and April of 2017, which was above the
30-year norm and had 98 and 38.8 mm greater precipitation relative to
the springs of 2015 and 2016, respectively.
Cumulative precipitation during the cash crop growing seasons of
2014, 2015, and 2016 hydrologic years was +152.9, +114.6, and
−89.7 mm compared to the 30-year average, respectively. In 2014
hydrologic year, which was the wettest of the three years; May, June
and July cumulative precipitation were 29.2 and 92.8 mm greater re-
lative to the same period in 2016 or 2017, respectively. Specifically,
June of 2015 had the highest precipitation on record for the state of
Illinois.
3.2. Cover crop biomass and N uptake
The main effect of N application timing did not significantly affect
cover crop biomass or N uptake. For this reason, results and discussion
regarding cover crop growth will use averages across N application
timing unless otherwise stated.
3.2.1. 2014–2015 cover crop growing season
During the fall of 2014, the rye/radish mixture accumulated a total
biomass of 298.7 kg ha
−1
and a total N uptake of 11.7 kg ha
−1
. By the
following spring, rye accumulated significantly greater biomass
(1106.7 kg ha
−1
), and N uptake (53.6 kg N ha
−1
) compared to fall of
2014 (Tables 3 and 4). The 2014–2015 cover crop growing season was
coldest observed (Fig. 1) during the study, which potentially limited fall
rye and radish growth. Cold conditions combined with below average
spring precipitation likely limited spring rye growth as well.
3.2.2. 2015–2016 cover crop growing season
The 2015 cover crop season preceded a soybean cash crop, and no N
fertilizer was applied. In contrast to the other CC seasons, the CC
mixture only had the potential to interact with naturally mineralized N
and residual N within the soil from the previous corn season. In the fall
of 2015, the CC mixture accumulated a total biomass of 1417.3 kg ha
−1
and absorbed 59.4 kg N ha
−1
. Above average fall temperatures and
precipitation provided an opportunity to accumulate CC growth. As a
result, significantly greater fall biomass and N uptake occurred relative
to the fall of 2014 and 2016. Specifically, above average air tempera-
tures in November 2015 contributed to increased CC growth and a late-
November radish termination date.
M.D. Ruffatti et al. Agricultural Water Management 211 (2019) 81–88
83
By chemical termination in the spring of 2016, rye accumulated
1223.3 kg ha
−1
of shoot biomass and 31.4 kg N ha
−1
of N uptake;
which was on average less than the total growth that occurred in the
preceding fall. However, spring rye resulted in significantly greater
biomass ( + 587.4 kg ha
−1
) and greater N uptake ( + 6.3 kg N ha
−1
)
relative to the rye alone growth observed in the previous fall.
3.2.3. 2016–2017 cover crop growing season
In the fall of 2016, the CC mixture accumulated an average total
biomass of 617.7 kg ha
−1
and absorbed 22.9 kg N ha
−1
. This fall was
the driest of the study with precipitation 99.3 mm below the 30-year
average. It is likely, that dry conditions limited both rye and radish fall
growth. In contrast to the fall of 2015, in which radish accounted for
the largest percentage of fall growth; fall 2016 radish growth resulted in
only 135.7 kg ha
−1
of shoot biomass and 5.3 kg N ha
−1
of N Uptake.
Table 2
Average ambient air temperature and annual precipitation for 2014, 2015, 2016 hydrologic years.
Temperature, °C Precipitation, mm
2014 2015 2016 30-yr Avg 2014 2015 2016 30-yr Avg
September 17.7 20.3 20.5 18.8 98.8 69.1 78.5 83.4
October 11.3 12.2 14.6 12.0 104.1 45.7 42.9 86.1
November 0.6 7 7.3 4.9 41.9 100.1 66 78.2
December −0.1 4.2 −2.4 −1.8 201 151.6 21.6 60.6
January −4.6 −3.6 −1−3.8 39.9 15.7 37.6 57.5
February −8.3 −0.4 3.9 −2.1 13.7 19.1 13 51.8
March 2.5 7.7 4.7 4.3 22.4 74.7 85.1 63.3
April 11.4 10.5 13.1 10.9 60.2 67.1 95.5 90.7
May 18 16.6 16 17.1 131.6 102.9 73.9 108.1
June 21.5 23.2 22.6 22.2 179.1 102.4 95 100.5
July 22.3 23.2 23.15 23.9 139.2 157 32.8 98.3
August 21.2 23.2 19.83 22.9 104.1 153.4 109.8 94.2
Total Average 9.5 12.0 11.9 10.8 1136.0 1058.8 751.7 972.7
Deviation from 30-yr Avg −1.3 1.2 1.1 163.3 86.1 −221.0
Fig. 1. Deviation from the 30-year regional average precipitation versus de-
viation from the 30-year regional average temperature for the 2014, 2015, and
2016 hydrologic years separated by cover crop and cash crop season.
Table 3
Cover crop shoot biomass for each species at both the fall and spring sampling dates for the 2014–2015, 2015–2016, and 2016–2017 years.
Sampling Period
2014–2015 2015–2016 2016–2017
Cover Crop Fall
b
Spring Fall Spring Fall Spring
Treatment Species kg ha
−1
FCC
a
Radish 755.9 126.5
Cereal Rye 332.2
Ac
1179.6
B
619.5
A
1072.7
B
422.7
A
2283.5
B
SCC Radish 806.9 144.8
Cereal Rye 265.2
A
1033.7
B
652.2
A
1373.8
B
541.3
A
2110.5
B
a
FCC = fall applied nitrogen with cover crops, SCC = spring applied nitrogen with cover crops.
b
Species were not separated in the fall of 2014, thus the value represented in the fall 2014 column is representative of both cereal rye and daikon radish biomass
accumulation.
c
Different capital letters indicate significant differences within species between the fall and spring sampling dates for a given year at an alpha level of 0.05.
Table 4
Cover crop nitrogen uptake for each species at both the fall and spring sampling
dates for the 2014–2015, 2015–2016, and 2016–2017 years.
Sampling Period
2014–2015 2015–2016 2016–2017
Cover Crop Fall
‡
Spring Fall Spring Fall Spring
Treatment Species kg N ha
−1
FCC
†
Radish 32.2
a
5.1
a
Cereal Rye 12.3
Aa§
61.5
Ba
22.7
Aa
29.1
Aa
15.9
Aa
91.5
Ba
SCC Radish 36.4
a
5.5
a
Cereal Rye 11.0
Aa
45.6
Ba
27.5
Aa
33.7
Aa
19.2
Aa
70.3
Bb
†
FCC = fall applied nitrogen with cover crops, SCC = spring applied ni-
trogen with cover crops.
‡
Species were not separated in the fall of 2014 and thus, the value re-
presented in the fall 2014 column is representative of both cereal rye and
daikon radish biomass accumulation.
§
Different capital letters indicate significant differences within treatment
and year among sampling periods at an alpha level of 0.05. Different lower case
letters within the same sampling period and cover crop species indicate sig-
nificant differences among treatments at an alpha level of 0.05.
M.D. Ruffatti et al. Agricultural Water Management 211 (2019) 81–88
84
Rye, however, accounted for 78% (482.0 kg ha
−1
) and 77% (17.6 kg N
ha
−1
) of the total fall shoot biomass and N uptake, respectively. Above
average temperatures in February and March 2017 allowed the rye to
begin growth earlier, relative to the first two springs. This combined
with greater precipitation, relative to spring 2015 or 2016, resulted in
ideal rye growing conditions. As a result, the greatest rye biomass
(2197 kg ha
−1
) and N uptake (80.9 kg N ha
−1
) was observed in the
spring of 2017.
3.3. Corn and soybean yields
In 2015, the inclusion of CC significantly reduced corn grain yields
by 7% within the spring dominated nitrogen application system.
However, CC did not significantly affect corn grain yields in the fall
dominated nitrogen application system. Furthermore, nitrogen appli-
cation timing had no significant effect on corn grain yields (Table 5).
During the 2016 soybean year, no significant differences were observed
in grain yield between any of the experimental treatments. The pre-
sence of CC resulted in significant decreases in corn grain yield in both
the fall (22%) and spring (16%) dominated nitrogen application sys-
tems in 2017. However, like the previous corn season, no significant
differences were observed as a result of N application timing (Table 5).
3.4. Nitrogen application timing and cover crop inclusion on water quality
3.4.1. Impact of nitrogen application timing on tile Nitrate-N loss
In the 2014 hydrologic year, tile water flow-weighted NO
3
-N con-
centration was significantly greater in the fall dominated nitrogen
treatment relative to the spring dominated nitrogen treatment by 18%
(Table 6); though this did not translate to significantly different NO
3
-N
loads between the two application timings (Table 7). However, in the
2015 or 2016 hydrologic years, nitrogen application timing did not
have a significant impact on annual average flow-weighted NO
3
-N
concentrations. Likewise, there was not a significant impact on annual
tile N loads during any year of the study.
Soybeans were grown during the 2015 hydrologic year as part of the
on-going corn-soybean crop rotation common to the region; therefore,
no N fertilizer was applied. Though no N fertilizer was applied, there
were slightly greater concentrations in the spring dominated nitrogen
treatment relative to the fall dominated nitrogen treatment; which
agrees with the 1.65 times greater NO
3
-N load in the spring dominated
nitrogen treatment compared to the fall dominated nitrogen treatment.
These findings are similar to those of Pittelkow et al. (2017) who re-
ported on average greater NO
3
-N concentration and load in the soybean
phase of a corn-soybean rotation when N was applied in the spring
relative to the fall.
During the 2016 hydrologic year, the fall dominated nitrogen
treatment resulted in 18% greater NO
3
-N flow-weighted concentration
relative to the spring dominated nitrogen treatment; however, NO
3
-N
loads were 14% lower in the fall dominated nitrogen treatment com-
pared to the spring dominated nitrogen treatment. Though not sig-
nificant, three-year average annual NO
3
-N concentrations were 11%
less in the spring dominated nitrogen treatment compared to the fall
dominated nitrogen treatment; but NO
3
-N loads were 25% greater in
the spring dominated nitrogen treatment (48.9 kg N ha
−1
) relative to
the fall dominated nitrogen treatment (39.0 kg N ha
−1
)(Tables 6 and
7).
3.4.2. Impact of cover crops on tile Nitrate-N loss
Cover crops did not significantly affect flow-weighted NO
3
-N con-
centration or load during the 2014 hydrologic year within N application
timing. Averaged across N timing, CC inclusion resulted in 10% less
flow-weighted NO
3
-N concentration and 11% less NO
3
-N load. The
spring dominated nitrogen with cover crop treatment significantly re-
duced the NO
3
-N concentration by 24% compared to the fall dominated
nitrogen treatment and resulted in an N load reduction of 34% (Tables 6
and 7, and Fig. 2a). There were no significant differences in flow-
weighted concentration or N load for the fall dominated nitrogen with
cover crop treatment relative to the spring dominated nitrogen treat-
ment.
During the 2015 hydrologic year, cover cropping significantly re-
duced NO
3
-N concentration by 39% and NO
3
-N load by 59% in the
spring dominated nitrogen with cover crop treatment relative to the
spring dominated nitrogen treatment. Furthermore, there was a trend
for cover crops to reduce flow-weighted NO
3
-N concentration
(P = 0.119) 35% and NO3-N load (P = 0.187) by 50% in the fall
dominated nitrogen with cover crop treatment compared to the fall
dominated nitrogen treatment. No significant difference between the
spring dominated nitrogen with cover crop and fall dominated nitrogen
treatments were observed. However, the flow-weighted NO
3
-N and N
loads were significantly less in the fall dominated nitrogen with cover
crop treatment relative to spring dominated nitrogen treatment by 40%
Table 5
Grain yields for the 2015, 2016, and 2017 cash crop growing seasons.
2015 Corn 2016 Soybeans 2017 Corn
Treatment
a
Mg ha
−1
FN 12.76
ABb
3.96
A
13.77
A
FCC 12.74
AB
3.77
A
10.69
B
SN 13.19
A
4.07
A
13.91
A
SCC 12.28
B
3.90
A
11.41
B
ZC 4.59
C
3.97
A
5.32
C
a
FN = Fall Nitrogen, FCC = Fall Nitrogen with Cover Crop, SN = Spring
Nitrogen, SCC = Spring Nitrogen with Cover Crop, ZC = Zero Control.
b
Different capital letters within a column denote significant differences
between treatments within year at an alpha level of 0.05.
Table 6
Average annual flow-weighted NO
3
-N concentrations for each treatment for the
2014–2016 hydrologic years.
2014 Hydrologic
Year
2015 Hydrologic
Year
2016 Hydrologic
Year
Treatment
a
mg L
−1
FN 10.48
Bb
5.44
AB
13.98
A
FCC 9.52
AB
3.53
A
5.48
B
SN 8.88
AC
5.97
A
11.80
A
SCC 7.93
AC
3.64
B
4.65
B
ZC 7.18
C
4.29
AB
8.92
C
a
FN = Fall Nitrogen, FCC = Fall Nitrogen with Cover Crop, SN = Spring
Nitrogen, SCC = Spring Nitrogen with Cover Crop, ZC = Zero Control.
b
Different capital letters within a column indicate significant differences
between experimental treatments for that given hydrologic year at an alpha
level of 0.05.
Table 7
Average annual total NO
3
-N load for each treatment for the 2014–2016 hy-
drologic years.
2014 Hydrologic
Year
2015 Hydrologic
Year
2016 Hydrologic
Year
Treatment
a
kg ha
−1
FN 42.45
Ab
38.46
AB
36.18
A
FCC 37.51
A
19.35
A
13.81
B
SN 40.73
A
63.83
B
42.32
A
SCC 36.04
A
25.85
A
15.12
B
ZC 44.71
A
58.29
AB
42.32
AB
a
FN = Fall Nitrogen, FCC = Fall Nitrogen with Cover Crop, SN = Spring
Nitrogen, SCC = Spring Nitrogen with Cover Crop, ZC = Zero Control.
b
Different capital letters within a column indicate significant differences
between experimental treatments for that given hydrologic year at an alpha
level of 0.05.
M.D. Ruffatti et al. Agricultural Water Management 211 (2019) 81–88
85
and 70%, respectively.
In the 2016 hydrologic year, both cover crop treatments resulted in
a significant reduction of 61% in flow-weighted NO
3
-N concentration
when compared to their respective non-cover cropped treatments. CC
inclusion resulted in significant NO
3
-N load reductions of 62% within
the fall N system and 64% for the spring N system. The spring domi-
nated nitrogen with cover crop treatment significantly reduced both
NO
3
-N concentration and N load compared to the fall dominated ni-
trogen treatment by 67% and 58%, respectively. Additionally, both
NO
3
-N concentration and load were 54% and 67% less in the fall
dominated nitrogen with cover crop treatment relative to the spring
dominated nitrogen treatment.
3.4.3. Tile Nitrate-N loss from the zero control treatment
Within the 2014 hydrologic year, the tile flow-weighted NO
3
-N
concentration (7.18 mg L
−1
) of the zero control was significantly less
relative to the fall dominated nitrogen (10.48 mg L
−1
) and fall domi-
nated nitrogen with cover crop (9.52 mg L
−1
) treatments and was si-
milar to both the spring dominated nitrogen and spring dominated ni-
trogen with cover crop treatments (Table 6). The zero control flow-
weighted NO
3
-N concentration (4.29 mg L
−1
) was similar to all treat-
ments in the 2015 hydrologic year. During the 2016 hydrologic year,
flow-weighted NO
3
-N concentrations in the fall dominated nitrogen
with cover crop (5.48 mg L
−1
) and spring dominated nitrogen with
cover crop (4.65 mg L
−1
) treatments were significantly lower relative to
the zero control (8.29 mg L
−1
). On the other hand, the fall dominated
nitrogen (13.98 mg L
−1
) and spring dominated nitrogen (11.80 mg L
−1
)
treatments resulted in significantly greater flow-weighted NO
3
-N con-
centrations when compared to the zero control (8.92 mg L
−1
). When
considering NO
3
-N load, there was no significant difference between
the zero control and all other treatments that received fertilizer N over
the course of the experiment. However, on average the NO
3
-N loading
for treatments augmented with CC was 66% lower than the zero control
(Table 7).
4. Discussion
4.1. Cover crop biomass growth and N uptake
Cover crops have been identified within the NLRS of UMRB states as
the most efficient in-field practice to mitigate non-point source NO
3
-N
contributions to surface waterways. The efficacy of cover crops to
generate NO
3
-N reduction depends primarily on cover crop growth. In
the current study, we observed considerable year to year variation in
radish shoot biomass (120–894.6 kg ha
−1
) and N uptake (5.7–40.6 kg N
ha
−1
); however, fall rye shoot biomass (465.2–683.3 kg ha
−1
) and N
uptake (17.2–26.1 kg N ha
−1
) was relatively consistent. These ob-
servations suggest that radish influenced the variation in fall growth of
the cover crop mixture. The data also demonstrated that the selection of
a cover crop mixture allows for responsive cover crop growth in the
event of abnormal increases in air temperature and precipitation in the
fall that create ideal conditions for N losses via leaching through tile
Fig. 2. Monthly average flow-weighted NO
3
-N concentrations and precipitation by rain fall event. for September 2014 through June 2017. Significant differences at
an alpha level of 0.05 between treatments within a given month are denoted with an asterisk (*). FN = Fall Nitrogen, FCC = Fall Nitrogen with Cover Crop,
SN = Spring Nitrogen, SCC =Spring Nitrogen with Cover Crop.
M.D. Ruffatti et al. Agricultural Water Management 211 (2019) 81–88
86
drainage. Spring rye biomass (1098.7–2208.3 kg ha
−1
) and N uptake
(50.6–83.3 kg N ha
−1
) were similar to what has been reported in the
literature for rye grown as a monoculture (Johnson et al., 1998;Dean
and Weil, 2009;Kaspar et al., 2012;Lacey and Armstrong, 2015). White
and Weil (2010) also examined a rye/radish mixture over a two-year
study. They reported rye, radish, and combined biomass ranges
763–869 (kg ha
−1
), 771–1821 (kg ha
−1
), and 1640–2584 (kg ha
−1
),
respectively. Observed rye and radish biomass ranges reported in that
study differed from our study likely due to variation in location,
weather, cash crop N management, and cover crop management.
4.2. Impact of nitrogen application timing and zero control on tile NO
3
-N
losses
In the UMRB, a consistent emphasis has been placed on transi-
tioning fall-applied N to the spring as pre-plant or side-dress applica-
tions closer to the timing of peak N demand of corn to reduce the
susceptibility of N loss from agricultural fields (Illinois Nutrient Loss
Reduction Strategy, 2015;Ribaudo et al., 2012). In our study, this
transitioning of fall-applied N to the spring resulted in an 11% reduc-
tion in flow-weighted NO
3
-N and a 25% increase in NO
3
-N load over a
three-year period. This trend of variability in NO
3
-N load reductions
between fall and spring N management is consistent with the literature
that suggests switching from a fall N system to a majority spring N
system results in a −67 to 52% reduction in NO
3
-N load via tile drai-
nage, with an average of 9.3% (Randall et al., 2003;Randall and
Vetsch, 2005;Randall and Mulla, 2001;Dinnes, 2004;Rejesus and
Hornbaker, 1999). Specifically, in our study, when considering NO
3
-N
load, we observed 66% greater NO
3
-N load for the spring dominated
nitrogen treatment compared to the fall dominated nitrogen treatment
during the soybean. This trend is corroborated by other tile-drainage
studies within the UMRB that found 32% less and 88% greater N load
for spring N versus fall N during soybean years (Randall and Vetsch,
2005;Pittelkow et al., 2017). These observations indicate that in spring
N systems, reductions in excessive NO
3
-N leaching could be attributed
to a timely N application and corn N uptake. However, in the sub-
sequent fallow period and soybean growing season a significant mass of
residual N loss occurs via tile drainage when N fertilizer is applied in
the spring (Randall and Vetsch, 2005;Pittelkow et al., 2017). Ad-
ditionally, our data agree with the literature and confirms that
switching N application timing alone is not adequate to achieve N load
reductions in the UMRB that could lead to a significant reduction in the
size of the Gulf of Mexico Hypoxic Zone.
One of the unexpected observations from our study was the fact that
NO
3
-N loading via tile-drainage was similar for the zero control that did
not receive N fertilizer to both the fall dominated nitrogen and spring
dominated nitrogen that received the full rate of N fertilizer.
Contributing factors to this finding may be differences in corn yield and
N uptake between the zero control and fertilized treatments. Fertilized
plots resulted in an average of 2.5 times greater corn yield then the zero
control (Table 5). It is likely that unfertilized corn plants resulted in less
evapotranspiration and exerted less physical demand on the NO
3
-N in
the soil solution due to poor N nutrition and root development. Fur-
thermore, the soil at the experimental site had an average soil organic
matter of 3.4%, which could be contributing to increased inorganic N
within the soil solution.
4.3. Impact of cover crop inclusion on spring and fall N application systems
In our study, combining a spring N application with cover crops
resulted in a 39% and 47% reduction in NO
3
-N flow-weighted con-
centration and load, respectively. These reductions were consistent
with what was observed in the literature, where cover crops in spring N
application systems reduced flow-weighted NO
3
-N concentration by
30–59% (Kaspar et al., 2007;Kaspar et al., 2012;Drury et al., 2014)
and NO
3
-N load by 12–61% (Strock et al., 2004;Kaspar et al., 2007,
2012;Drury et al., 2014). With the exception of Strock et al. (2004),
who reported that rye reduced flow-weighted NO
3
-N concentrations by
13% over a 3-year period, which was drastically lower due to the ap-
plication of cover crops only after corn.
To address N loading from Fall N application, which is common in
the UMRB, one objective our study was to compare the effectiveness of
augmenting a fall N application system with cover crops to a fall N
system without cover crops. We observed that CC reduced the mean
annual flow-weighted NO
3
-N concentrations of tile water by 38% and
NO
3
-N load by 40% when added to a fall N application system. To the
authors’knowledge, the data presented in this study represent the only
report of the impact of CC on tile NO
3
-N in fall N application systems.
However, these reductions are similar to those in spring N application
systems reported in both this study and the literature. Suggesting that,
despite N application timing, cover crops are effective at reducing tile N
losses.
Additionally, we found that NO
3
-N reductions, due to cover crops,
occurred across the entire hydrologic year (Fig. 2b and c), demon-
strating that the impact of CC on water quality is not limited only to the
CC growing season. Moreover, unlike N application timing, CC inclu-
sion resulted in water quality benefits in years where N was not applied.
For example, we observed an average reduction of 37% in the soybean
year, which is similar to 42% reported by Kaspar et al. (2007). This
annual impact of CC on N loading could be attributed N cycling, where
CC are altering the concentration and distribution of available N in the
soil profile making a smaller mass of NO
3
-N susceptible to loss through
tile drainage (Lacey and Armstrong, 2014;Johnson et al., 1998). In our
study, an example of N cycling is that on average we found a 2:1 ratio of
CC shoot N relative to N that cover crops prevented from leaving via
tile-drainage. This is a conservative measure of CC N cycling potential
because it is not considering the N that is in the CC roots. Moreover,
both CC treatments reduced N loading by 66% on average relative to
the zero control. This indicates that CC are potentially absorbing
naturally mineralized N that contributes to N loss via tile-drainage.
Adjusting the N application timing from fall to spring application of
N and cover cropping have both demonstrated the potential to be ef-
fective N loss reduction methods. Combination the two methods re-
duced flow-weighted NO
3
-N concentrations by 39%. However, the
spring dominated nitrogen with cover crop treatment only resulted in
an average reduction of 46% in NO
3
-N concentration relative to the fall
dominated nitrogen treatment (Fig. 2d). This suggests that, while not
additive, the combination of spring N application timing and cover
cropping does result in increased efficacy to reduce NO
3
-N losses
through agricultural subsurface drainage systems.
Despite the long-standing effort to encourage farmer adoption of
spring N application systems in the UMRB, there are many watersheds
in the region where it is a common practice for a significant mass of N
to be fall applied as anhydrous ammonia, ammonium phosphate, and
manure (Bierman et al., 2012;Illinois Nutrient Loss Survey Results,
2016;Lemke et al., 2011;Ribaudo et al., 2012;Smiciklas et al., 2008).
Results of this study demonstrated that applying fall N into a living CC
stand resulted in a 30% and 52% reduction in NO
3
-N concentration and
load relative to applying N in the spring without the presence of CC.
This observation suggests that augmenting a fall N system with CC
could be an effective adaptive BMP that reduces the susceptibility of N
loading via tile drainage in watersheds dominated by fall-applied N.
5. Conclusion
The results of this study confirmed that right timing of N application
alone results in a variable impact on N loading via tile drainage and that
coupling in-field N reduction practices are most effective. Adoption of a
CC mixture allows for responsive cover crop growth in the event of
abnormal increases in air temperature and precipitation in the fall or
spring that results in ideal conditions for N losses through tile drainage.
Zero application of fertilizer N resulted in similar NO
3
-N loading
M.D. Ruffatti et al. Agricultural Water Management 211 (2019) 81–88
87
compared to fall and spring N management systems without cover
crops; however, augmenting those systems with cover crops drastically
reduced N loss when compared to the non-fertilized plots. The presence
of cover crops reduced corn yields but did not affect soybean yields. The
inclusion of CC was an effective NO
3
-N loss mitigation strategy, re-
gardless of the N fertilizer application timing. Finally, CC proved to be
effective in reducing NO3-N loading through tile-drainage across the
spectrum of N fertilizer management systems common to the UMRB.
Funding
This work was supported by the Illinois Nutrient Research and
Education Council [grant number 11580-02-502009170].
Acknowledgements
The authors would like to thank the collaborating producer for
performing all in-field applications. Also the authors acknowledge
Travis Deppe, Felix Vogel, Victoria Heath, and Clayton Nevins for their
time in sample collection, preparation and analysis.
References
Basche, A.D., Kaspar, T.C., Archontoulis, S.V., Jaynes, D.B., Sauer, T.J., Parkin, T.B.,
Miguez, F.E., 2016. Soil water improvements with the long-term use of a winter rye
cover crop. Agric. Water Manag. 172, 40–50.
Bierman, P.M., Rosen, C.J., Venterea, R.T., Lamb, J.A., 2012. Survey of nitrogen fertilizer
use on corn in Minnesota. Agric. Syst. 109, 43–52.
Blesh, J., Drinkwater, L.E., 2014. Retention of N-labeled fertilizer in an Illinois prairie soil
with winter rye. Soil Sci. Soc. Am. J. 78 (2), 496–508.
Cooper, R.J., Aziza, J.H., Hiscocka, K.M., Lovetta, A.A., Dugdalea, S.J., Sünnenberga, G.,
Noble, L., Beamish, J., Hovesen, P., 2017. Assessing the farm-scale impacts of cover
crops and non-inversion tillage regimes on nutrient losses from an arable catchment.
Agric. Ecosyst. Environ. 237, 181–193.
Daigh, A.L., Zhou, X., Helmers, M.J., Pederson, C.H., Ewing, R., Horton, R., 2014.
Subsurface drainage flow and soil water dynamics of reconstructed prairies and corn
rotations for biofuel production. Vadose Zone J. 13, 1–11.
David, M., McIsaac, G., Czapar, G., Schnitkey, G., Mitchell, C., 2013. Science Assessment
to Support an Illinois Nutrient Loss Reduction Strategy.
Dean, J.E., Weil, R.R., 2009. Brassica cover crops for nitrogen retention in the Mid-
Atlantic Coastal Plain. J. Environ. Qual. 38 (2), 520–528.
Dinnes, D.L., 2004. Assessments of Practices to Reduce Nitrogen and Phosphorus
Nonpoint Source Pollution of Iowa’s Surface Waters. USDA-ARS, National Soil Tilth
Laboratory, pp. 366.
Drury, C.F., Tan, C.S., Welacky, T.W., Reynolds, W.D., Zhang, T.Q., Oloya, T.O.,
McLaughlin, N.B., Gaynor, J.D., 2014. Reducing nitrate loss in tile drainage water
with cover crops and water-table management systems. J. Environ. Qual. 43 (2),
587–598.
Gardner, J.B., Drinkwater, L.E., 2009. The fate of nitrogen in grain cropping systems: a
meta-analysis of 15N field experiments. Ecol. Appl. 19 (8), 2167–2184.
Illinois Nutrient Loss Reduction Strategy, 2015. Illinois Environmental Protection Agency.
Accessed 12 Dec. 2017. http://www.epa.illinois.gov/topics/water-quality/
watershed-management/excess-nutrients/nutrient-loss-reduction-strategy/index.
Illinois Nutrient Loss Survey Results, 2016. National Agricultural Statistics Service.
Accessed 15 Jan. 2018. https://www.nass.usda.gov/Statistics_by_State/Illinois/
Publications/Current_News_Release/2016/Nutrient_Loss_Survey_Results.pdf.
Iowa Nutrient Reduction Strategy, 2013. A Science and Technology-based Framework to
Assess and Reduce Nutrients to Iowa Waters and the Gulf of Mexico. Iowa
Department of Agriculture and Land Stewardship. Iowa Department of Natural
Resources, and Iowa State University College of Agriculture and Life Sciences,
Ames, IA.
Johnson, T.J., Kaspar, T.C., Kohler, K.A., Corak, S.J., Logsdon, S.D., 1998. Oat and rye
overseeded into soybean as fall cover crops in the upper Midwest. J. Soil Water
Conserv. 53 (3), 276–279.
Kaspar, T.C., Jaynes, D.B., Parkin, T.B., Moorman, T.B., 2007. Rye cover crop and
gamagrass strip effects on NO concentration and load in tile drainage. J. Environ.
Qual. 36 (5), 1503–1511.
Kaspar, T.C., Jaynes, D.B., Parkin, T.B., Moorman, T.B., Singer, J.W., 2012. Effectiveness
of oat and rye cover crops in reducing nitrate losses in drainage water. Agric. Water
Manag. 110, 25–33.
Lacey, C.G., Armstrong, S.D., 2014. In field measurements of nitrogen mineralization
following fall applications of N and the termination of winter cover crops. Air Soil
Water Res. 7, 53.
Lacey, C., Armstrong, S., 2015. The efficacy of winter cover crops to stabilize soil in-
organic nitrogen after fall-applied anhydrous ammonia. J. Environ. Qual. 44 (2),
442–448.
Lemke, A.M., Kirkham, K.G., Lindenbaum, T.T., Herbert, M.E., Tear, T.H., Perry, W.L.,
Herkert, J.R., 2011. Evaluating agricultural best management practices in tile-
drained subwatersheds of the Mackinaw River, Illinois. J. Environ. Qual. 40 (4),
1215–1228.
Minnesota Nutrient Reduction Strategy, 2014. Minnesota Pollution Control Agency.
Accessed 10 Jan. 2018. https://elibrary.asabe.org/azdez.asp?search=0&JID=1&
AID=47346&CID=ids2016&v=&i=&T=2.
Mississippi River/Gulf of Mexico Watershed Nutrient Task Force, 2008. Gulf Hypoxia
Action Plan 2008 for Reducing, Mitigating, and Controlling Hypoxia in the Northern
Gulf of Mexico and Improving Water Quality in the Mississippi River Basin.
Washington, DC. .
Missouri Nutrient Loss Reduction Strategy, 2014. Missouri Department of Natural
Resources. Accessed 10 Jan. 2018. https://dnr.mo.gov/env/wpp/mnrsc/docs/nlrs-
strategy-2014.pdf.
Pittelkow, C.M., Clover, M.W., Hoeft, R.G., Nafziger, E.D., Warren, J.J., Gonzini, L.C.,
Greer, K.D., 2017. Tile drainage nitrate losses and corn yield response to fall and
spring nitrogen management. J. Environ. Qual. 46 (5), 1057–1064.
Qi, Z., Helmers, M.J., 2010. Soil water dynamics under winter rye cover crop in central
Iowa. Vadose Zone J. 9 (1), 53–60.
Randall, G.W., Mulla, D.J., 2001. Nitrate nitrogen in surface waters as influenced by
climatic conditions and agricultural practices. J. Environ. Qual. 30 (2), 337–344.
Randall, G.W., Vetsch, J.A., 2005. Nitrate losses in subsurface drainage from a corn–-
soybean rotation as affected by fall and spring application of nitrogen and nitrapyrin.
J. Environ. Qual. 34 (2), 590–597.
Randall, G.W., Vetsch, J.A., Huffman, J.R., 2003. Nitrate losses in subsurface drainage
from a corn–soybean rotation as affected by time of nitrogen application and use of
nitrapyrin. J. Environ. Qual. 32 (5), 1764–1772.
Rejesus, R.M., Hornbaker, R.H., 1999. Economic and environmental evaluation of alter-
native pollution-reducing nitrogen management practices in central Illinois. Agric.
Ecosyst. Environ. 75 (1), 41–53.
Ribaudo, M., Livingston, M., Williamson, J., 2012. Nitrogen Management on Us Corn
Acres, 2001–10. United States Department of Agriculture, Economic Research
Service.
Smiciklas, K.D., Moore, A.S., Adams, J.C., 2008. Fertilizer nitrogen practices and nitrate
levels in surface water within an Illinois watershed. J. Nat. Resour. Life Sci. Educ. 37
(1), 14–19.
Smil, V., 1999. Nitrogen in crop production: an account of global flows. Global
Biogeochem. Cycles 13 (2), 647–662.
Strock, J.S., Porter, P.M., Russelle, M.P., 2004. Cover cropping to reduce nitrate loss
through subsurface drainage in the northern US Corn Belt. J. Environ. Qual. 33 (3),
1010–1016.
White, C.M., Weil, R.R., 2010. Forage radish and cereal rye cover crop effects on my-
corrhizal fungus colonization of maize roots. Plant Soil 328 (1-2), 507–521.
Wisconsin Nutrient Reduction Strategy, 2013. Wisconsin Department of Natural
Resources. Accessed 10 Jan. 2018. https://dnr.wi.gov/topic/SurfaceWater/
nutrient/combined_draft.pdf.
M.D. Ruffatti et al. Agricultural Water Management 211 (2019) 81–88
88
