Greenhouse Gas Emissions Dynamics as Influenced by Corn Residue

0 downloads 0 Views 1MB Size Report
Mar 13, 2015 - Corn (Zea mays L.) residue is a potential feedstock source for ethanol ... Additionally, agriculture accounts for 10 to 20% of the to- ... ization rates, although can be reduced with residue removal. .... gas samples (10 mL) were collected with polypropylene syringes ... and the ideal gas laws (Ussiri et al., 2009).
Published March 13, 2015

Soil & Water Management & Conservation

Greenhouse Gas Emissions Dynamics as Influenced by Corn Residue Removal in Continuous Corn System Jose Guzman*

Ohio State Univ. Carbon Management and Sequestration Center Columbus, OH 43210

Mahdi Al-Kaisi

Iowa State Univ. Agronomy, Ames, IA 50011

Timothy Parkin

USDA-ARS and National Laboratory for Agriculture and the Environment Ames, IA 50011

The removal of corn residue for bioethanol may require changes in current tillage and fertilization practices to minimize potential alterations to the soil environment that may lead to increase in greenhouse gas (GHG) emission. The objectives of this study were to examine how tillage, N fertilization rates, residue removal, and their interactions affect CO2, and N2O soil surface emissions. Greater CO2 emission coincided with higher soil temperatures typically observed with conventional tillage (CT) compared with no-tillage (NT), resulting in greater annual cumulative CO2 emission in CT (18.1 CO2 Mg ha−1 yr−1) compared with NT (16.2 CO2 Mg ha−1 yr−1) in 2009 and 2010 across sites. However, drier soil conditions during the growing season in 2011 lead to higher soil temperatures compared with 2009 and 2010. Consequently, annual cumulative CO2 emission from NT with 50 and 100% residue removal was (19.5 CO2 Mg ha−1 yr−1) greater than that from CT (17.8 CO2 Mg ha−1 yr−1) across all residue removal rates and from NT (17.5 CO2 Mg ha−1 yr−1) with no residue removal, respectively across all N rates in the Ames central site (AC) in 2011. In the Armstrong southwest site (ASW) site, there were no significant differences between tillage or residue removal rates for annual cumulative CO2 emission (19.9 CO2 Mg ha−1 yr−1) in 2011. Although N2O emission was considerably lower than CO2 emission, differences in N fertilization rates did have a significant impact on global warming potential once these gases were converted on the basis of their radiative forcing of the atmosphere. Abbreviations: AC, Ames central site; ASW, Armstrong southwest site; GHG, greenhouse gas; GWP, global warming potential; NT, no-till; PVC, polyvinyl chloride; SOC, soil organic carbon; TC, total carbon; UAN, urea-ammonium nitrate 32% N.

C

orn (Zea mays L.) residue is a potential feedstock source for ethanol production that may contribute to the reduction of fossil fuel use and net GHG emissions (Wilhelm et al., 2004; Graham et al., 2007). Although it is currently more expensive to produce ethanol from cellulosic plant materials than from grains, it is projected that improvements in technology and scale of production will decrease these costs (Foust et al., 2009). It is probable that cellulosic ethanol production will become a viable option and could create an annual market for crop residue from approximately 146 million Mg to 172 to 279 million Mg by the year 2017 (Downing et al., 2011). The removal of crop residue, however, may require changes in current tillage and fertilization practices to minimize potential alterations to the soil environment that may lead to increase in GHG emission. It is well documented through research over the past many decades that crop residues are critical for replenishing soil organic carbon (SOC) through conservation practices, which include miniSoil Sci. Soc. Am. J. 79:612–625 doi:10.2136/sssaj2014.07.0298 Accepted 3 Jan. 2015. Received 16 July 2014. *Corresponding author ([email protected]). © Soil Science Society of America, 5585 Guilford Rd., Madison WI 53711 USA 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. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.



Soil Science Society of America Journal

mum tillage, proper management of residues, and adequate soil fertility (Al-Kaisi and Yin, 2005; Al-Kaisi and Kwaw-Mensah, 2007). Crop residues also have significant roles in improving physical and chemical properties that are essential in protecting soil by controlling wind and water erosion, which ultimately reduce sediments and other contaminants transported to water bodies (Lindstrom, 1986; Karlen et al., 1994). Additionally, agriculture accounts for 10 to 20% of the total anthropogenic GHG emissions, but it is responsible for 25 and 58% of the total anthropogenic carbon dioxide (CO2) and nitrous oxide (N2O) emissions, respectively (Cole et al., 1997; Smith et al., 2007). Under aerobic soil conditions, agriculture soils typically are a minor emitter or small sink for methane CH4 (Bronson and Mosier, 1994). However, many strategies aimed for agriculture practices to mitigate GHG emissions have been proposed and investigated, including: crop rotations, changes in fertilizer regimes, and tillage while still increasing crop yields (Paustian et al., 2000; Johnson et al., 2007). Differences in type of tillage, nitrogen (N) fertilization, and residue management can have large effects on GHG emissions. Less intensive tillage such as NT that reduces soil disturbance and C as well as N mineralization can potentially lower CO2 and N2O emissions (Drury et al., 2006; Snyder et al., 2009). However, NT soils can also have greater N2O emission through denitrification processes than those under more intensive tillage (Burford et al., 1981; Linn and Doran, 1984; MacKenzie et al., 1997). This is particularly the case in NT soils that have a greater bulk density, resulting in reduced diffusion of gases in the soil, and increased water content creating anaerobic conditions favorable to denitrification processes (Mosier et al., 2002). Increases in N fertilization in most cases result in greater N2O emission (Bouwman, 1996; Pelster et al., 2011), although effects on CO2 emission are variable (Al-Kaisi et al., 2008). Little research has been conducted on residue management effects on CO2 and N2O emissions. Although, it is largely accepted that having greater amounts of crop residues in the soil will increase C and N mineralization, thus promoting CO2 and N2O production (Cochran et al., 1997; Paustian et al., 2000; Mosier et al., 2002). Since CO2, N2O, and CH4 emissions are mostly driven by biological processes, the rates of theses fluxes primarily depends on the availability of C substrate for CO2 and CH4 production and mineral N source for nitrification or denitrification, as well as soil temperature, soil water content, and oxygen availability for all the aforementioned GHGs (Linn and Doran, 1984; Cochran et al., 1997; Mosier et al., 2002). During tillage, the soil is loosened and mixed to incorporate crop residues and other amendments into the tillage zone in preparation for establishing a good seedbed for planting, resulting in warmer and drier soil conditions compared with NT (Licht and Al-Kaisi, 2005; Al-Kaisi and Kwaw-Mensah, 2007). This also accelerates the loss of soil organic matter as CO2 through oxidation and mineralization processes because of aerating and increasing soil temperature (Al-Kaisi and Yin, 2005). Nitrogen fertilization and residue management can also indirectly affect CO2 and N2O emissions www.soils.org/publications/sssaj

by altering soil temperature and soil water content by shading and water uptake by plants (Sainju et al., 2012). Unfortunately, there are very few comprehensive studies that have examined the soil, environmental, and management practice effects on CO2, and N2O soil surface emissions. We hypothesized that GHG emissions would be greater with management practices that have higher intensity of tillage and N fertilization rates, although can be reduced with residue removal. The objectives of this study were to examine how tillage, N fertilization rates, residue removal, and their interactions affect soil temperature, soil water content, and soil mineral N, and how these factors affect CO2 and N2O soil surface emissions in central and southwest Iowa. Additionally, this study quantified cumulative GHG emissions effects on global warming potential (GWP) in Central Iowa.

MATERIALS AND METHODS

Experimental Sites and Treatments A study was established in fall of 2008 on a Nicollet (fine-loamy, mixed superactive, mesic Aquic Hapludoll) and Canisteo clay loam (Fine-loamy, mixed, superactive, calcareous, mesic Typic Endoaquolls) soil association at Iowa State University Agronomy Research Farm (AC) in central Iowa (42°4¢ N, 95°5¢ W) and a Marshall silty clay loam (fine-silty, mixed, superactive, mesic Typic Hapludolls) soil association at Armstrong Research and Demonstration Farm (ASW) in southwest Iowa (41.3° N; 95.1° W) in continuous corn. The mean air temperature and annual precipitation at the AC site are 8.7°C and 975 mm, respectively (data from 1982 to 2011). At the ASW site, mean air temperature and annual precipitation are 9.5°C and 909 mm, respectively (data from 1982 to 2011). Both sites were previously in corn–soybean [Glycine max (L.) Merr.] rotation under CT, chisel plow in fall and chisel plus disk in the spring to a depth of 15 cm) and nonirrigated. The N application history for the sites was urea-ammonium nitrate 32% N (UAN) solution that was side-dressed injected in May after planting with agronomic rates of 170 kg N ha−1 during the corn phase (Blackmer et al., 1997). Phosphorus and potassium fertilization was used periodically to maintain optimum soil concentrations so as not to restrict corn growth during the length of this study. Treatments were established in 2008 to 2011 for both sites in a randomized, complete-block design with split-split arrangement and three replications. Each individual plot dimension was 6 m wide by 15 m long. The main treatment was tillage practice (NT and CT), which was split into three different residue removal rates (0, 50, and 100%), which was further split into three N fertilization rates of 0, 170, and 280 kg N ha−1. As done previously during past management, source of N was 32% liquid UAN, which was injected side-dressed in May after planting. Shortly after corn harvest, desired rates of residue removal were accomplished by adjusting down pressure on raking equipment before baling of residue. For 100% removal, corn stalks and leaves were first mowed then raked clear down (very high downpressure) for residue to be collected by baler. After baling, 613

plots were hand raked to achieve nearly 100% removal. For 50% removal, residue was not mowed and a decreased down pressure of the rake equipment was set so as to leave approximately 50% of the soil surface covered after baling. Actual removal of residue by mass varied by N rate, tillage, site, and year ranging from 46 to 64% and 90 to 99% for 50 and 100% residue removal treatments, respectively. Both sites were planted with a 111-d maturity corn variety (P33W84) at a seeding density of 79,000 seeds ha−1.

Greenhouse Gas Emissions, Soil Temperature, and Water Content Measurements During the growing season from April to October of each year, CO2 emission measurements were taken in 6- to 10-d intervals coupled with soil moisture (TRIME-FM Time Domain Reflectometry, Mesa, Medfield, MA) and temperature (thermometer attached to LI-COR 6400) at 5-cm soil depth in each plot with a portable infrared CO2 gas analyzer (LI-COR 6400, LI-COR, Lincoln, NE) with a soil respiration chamber. Measurements were taken between 0800 and 1100 h to approximate the 24 h mean soil surface CO2 emission. Soil surface CO2 measurements were conducted by placing a soil respiration chamber over a 10-cm-diameter polyvinyl chloride (PVC) ring placed into the ground for the entire year at 3-cm soil depth and leaving ~2 cm of the ring above the soil surface. Two PVC rings were placed in each plot, one within the plant row and one in between plant rows. The mean of the two rings was considered to be the soil surface CO2 emissions for the entire plot. During the nongrowing season, biweekly or monthly readings were taken because of lower GHG emissions and low soil temperatures. During the same day as CO2 emission was measured, soil N2O emission was measured following sampling protocol of GRACEnet chamber-based trace gas flux measurement (Parkin and Venterea, 2010). Two PVC rings (30 cm diameter and 10 cm high) were installed in each plot to a depth of approximately 6 cm. In each plot, one ring was placed directly in the plant row. The other ring was placed between plant rows on top of the UAN band. The N2O flux measurements were performed by placing vented chambers (30 cm diameter and 10 cm high) on the PVC rings and collecting gas samples at 0, 30, and 60 min following chamber deployment. At each time point, chamber headspace gas samples (10 mL) were collected with polypropylene syringes and immediately injected into evacuated glass vials (6 mL) fit with butyl rubber stoppers. Nitrous oxide concentrations in samples were determined with a gas chromatography instrument (Model GC17A; Shimadzu, Kyoto, Japan) equipped with a 63Ni electron capture detector and a stainless steel column (0.3 cm diameter and 74.54 cm long) with PorapakQ (80–100 mesh). Methane was analyzed with a flame ionization detector and a 0.3-cm-diameter and 90-cm-long column. Gas fluxes were calculated as changes in linear concentration gradient over time and the ideal gas laws (Ussiri et al., 2009). Cumulative soil surface CO2 emission for the growing season was calculated as follows (Grote and Al-Kaisi, 2007):

614

n

Cumulative N 2O or CO 2 ( kg ha −1 ) = ∑ i

( Xi +

Xi + 1) * (ti+1 -t1 ) 2

[1]

where, Xi is the first CO2 emission (kg ha−1 d−1) reading, and Xi+1 is the following reading at times ti and ti+1, respectively; n is the last CO2 emission reading during the growing season and i is the first CO2 emission reading in the growing season. Cumulative soil surface N2O emission was calculated as done in CO2 calculations. Since soil surface CO2 and N2O emissions were only measured weekly during the growing season, a linear regression model with soil surface temperature as the predictor variable was used to estimate missing weekly soil surface CO2 (R2 0.8) and N2O (R2 0.75) emission measurements (Fang and Moncrieff, 2001; Dornbush and Raich, 2006). This resulted in an additional cumulative 1500 kg CO2 ha−1 and 0.32 kg N2O ha−1 on average from early November to late March for the 2009 through 2011 seasons in both sites. During the November to March period, soil surface CO2 and N2O emissions across treatments were not significantly different and remained relatively low compared with growing season measurements because of soil being mostly near or below freezing temperatures (Fang and Moncrieff, 2001; Dornbush and Raich, 2006). Soil surface CO2 and N2O emissions during winter months represented approximately 9 and 6% of the total annual soil surface emissions, respectively, which is within the range reported in studies done in the north-central United States (Kessavalou et al., 1998; Dornbush and Raich, 2006).

Soil Mineral Nitrogen Measurements Mineral N (NO3–N and NH4–N) of soil samples were determined by KCl extraction method 3 d after N fertilization and approximately every 2 wk afterward until late September (Mulvaney, 1996). Ten to twelve soil cores at 1.7 cm diameter and 60-cm soil depth were collected. Approximately 10 g of airdried soil was placed in a 125-mL Nalgene bottle along with 50 mL of 2 M L−1 KCl. The Nalgene bottles were shaken for 30 min, and after that the extraction solution was filtered through Whatman No. 42 filter paper into 20-mL scintillation vials. The filtered solution was stored in a −4°C freezer until analyses for NO3–N and NH4–N was done with a Lachat QuickChem 8000FIA+ (Lachat Instruments, Milwaukee, WI).

Soil Organic Carbon Measurements Soil samples were collected after harvest in mid-October, before establishment of treatments in 2008 and 2 yr afterward in 2010, to measure changes in SOC. Twelve 1.7-cm-diameter soil cores were collected from depths of 0 to 15, 15 to 30, 30 to 45, and 45 to 60 cm in each treatment plot. Soil cores for each depth were combined into a single homogeneous sample. Soil samples at field moisture were 2 mm sieved and then air dried before being analyzed. The total carbon (TC) was determined with dry combustion by a CN analyzer (LECO Corporation, St. Joseph, MI). A separate carbonate analysis using a modified pressure calcimeter method was used to determine inorganic C concentraSoil Science Society of America Journal

tion, which was subtracted from TC concentration to calculate SOC concentration (Sherrod et al., 2002). Concurrently while TC soil samples were being taken, three individual bulk density (rb) samples were collected with a 1.7-cm-diameter soil probe for each depth from each treatment plot. Soil cores were taken at the same soil depths as TC samples and were then oven dried at 105°C for 24 h and weighed. Bulk density (Mg m–3) was calculated as the dried soil mass divided by the soil core volume. The SOC concentrations (mg g−1 dry soil) that were measured in 2008 to2010 were multiplied by mean rb values and soil depth thickness to convert SOC concentrations to mass per area basis (Mg ha−1) per plot basis for all treatment plots by soil depth. For soil samples collected after the baseline year in 2008, SOC content was calculated similar to initial samples, although concentrations of SOC for each soil profile depth were adjusted for gains or losses in soil mass because of changes in rb for the original or equivalent soil mass in 2008 to determine changes in SOC stocks (Lee et al., 2009).

squares fitting technique was used to assess the impact of soil temperature, soil water content, and mineral N on soil surface GHG emissions by multiple linear regression analysis in SAS.

Results

Soil Temperature, Water Content, and Mineral Nitrogen

The effects of residue removal on soil temperature, water content, and mineral N varied by site, year, tillage, and N fertilization. During the first year after residue removal in 2009, measurements were only taken in the AC site (Fig. 1). Soil temperatures were approximately 1°C warmer in NT compared with CT before the month of May when planting was done. After field cultivation and planting were completed and before the corn canopy completely covered the soil surface in early July, a tillage by residue removal interaction effect was observed for soil temperature. Soil temperature was 1°C cooler under no residue removal than when 50 and 100% of the residue was removed with NT. Under CT, there were no significant differences in soil Statistical Analyses temperature across the different residue removal rates, although Data for GHG emissions and soil temperature, water consoil temperature was on average 1°C warmer than NT when 50 tent, SOC, and mineral N were analyzed by using the analysis of and 100% of residue was removed and 1.5°C warmer when no repeated measures procedure in the Proc Mixed model of SAS residue was removed. Once the corn canopy had completely cov(SAS Institute, 2002). A compound symmetry covariance strucered the soil surface, there was no significant difference in tillage ture was used for repeated measures. Residue removal, tillage sysand residue removal effects on soil temperature until after grain tem, fertilizer N rate, and their interactions were considered fixed harvest and fall tillage (mid-October) was completed. During effects, while block and interactions with block were considered this time, CT was on average 1.2°C warmer than NT. Nitrogen random effects and date of measurement as the repeated measure fertilization also had a significant effect on soil temperature. variable. Mean soil surface GHG emissions between plant row and Beginning in early July, soil temperature was on average 0.5 to within plant row were used in this statistical analysis. Mean sepa1°C warmer when no N was applied compared with N applied ration was determined by the PDIFF procedure, and significance treatments. After late October, there were no significant differwas declared at P £ 0.10, unless otherwise stated. The linear least ences in soil temperature across tillage, residue removal rates, and N fertilization rate. Soil water content fluctuated with rainfall events throughout the season in the AC site in 2009 (Fig. 1). No-till on average had 0.32 m3 m−3 soil water content throughout the year and was 0.02 m3 m−3 higher than CT. There were no significant differences between residue removal rates or N fertilization rates on soil water content observed. For mineral N (NO3−–N and NH4+–N), average concentration in the soil before N was applied on 20 May was 0.3 mg N kg−1 soil (Fig. 2). Two days after N fertilization, mineral N significantly increased in NT on average to 153.6 mg N kg−1 soil. Under CT, mineral N varied by 150.6, 128.3, and 107.9 mg N kg−1 Fig. 1. Tillage system and residue removal rate effects on daily soil temperature and water content at time soil with 0, 50, and 100% residue of greenhouse gas measurement by nitrogen fertilization rate during the growing season at the Ames removal, respectively. Almost 3 wk central site (AC) in 2009. CT, conventional till; NT, no-till. www.soils.org/publications/sssaj

615

later, mineral N concentration stayed relatively the same in CT while under NT mineral N dropped on average by 50%. The following weeks up until mid-August, mineral N was greater under CT than in NT until it was depleted in the soil. There were no significant differences in mineral N across residue removal rates or tillage when no N was applied to the soil. In 2010, soil temperature in the AC site was on average 3.4°C warmer than in 2009 (18.1°C) during the growing season (Fig. 3). A parabolic trend in soil temperatures was observed as in 2009, peaking just before corn canopy completely covered the soil surface late in June; although, differences in soil temperature were only observed under different residue removal rates (Fig. 3). In treatments when no residue was removed, soil temperature was 0.5°C cooler than when 50 and 100% of the residue was removed from mid-May (after planting and field cultivation) until late June when corn canopy had mostly covered the soil surface. However, treatments in which N was not applied, the soil surface was still exposed to solar radiation throughout the growing season because of less can- Fig. 2. Tillage system and residue removal rate effects on soil mineral N by nitrogen opy shading, resulting in average of 0.8°C warmer fertilization rate during 2009 and 2010 growing seasons at AC. soil temperature than when N was applied to the up until late August when mineral N was depleted in the soil, soil across different tillage and residue removal rates. As in 2009, soil water content fluctuated with rainfall there were no differences in mineral N across residue removal events in the AC site for 2010 (Fig. 3). No-till on average had rates and tillage. In addition, there were no significant differences 0.35 m3 m−3 soil water content throughout the year, and it in mineral N across residue removal rates or tillage when no N was 0.03 m3 m−3 greater than CT. There were no significant was applied to the soil. differences across residue removal rates or N fertilization on In the ASW site, similar seasonal trends in soil temperatures soil water content observed. Before N was applied on 19 May, (Fig. 4) were observed as in the AC site although peaking later in soil mineral N concentration was 0.3 mg N kg−1soil (Fig. 2). Shortly after N fertilization (3 d), mineral N in the soil varied greatly across tillage and residue removal rates. The highest concentrations of mineral N occurred under NT when no residue was removed (420.0 mg N kg−1 soil), and when 50% of the residue was removed (304.8 mg N kg−1 soil). However, the lowest concentrations of mineral N were observed under NT when 100% of the residue was removed at 127.1 mg N kg−1 soil. Under CT, mineral N did not vary by residue removal rate, with an average of 193.3 mg N kg−1 soil. Twenty-four days later, mineral N in the soil significantly decreased under NT in contrast to CT where significant increases were observed. The following weeks Fig. 3. Tillage system and residue removal effects on daily soil temperature and water content at time of greenhouse gas measurement by nitrogen fertilization rate during the growing season at AC in 2010.

616

Soil Science Society of America Journal

Under no N fertilization, soil temperatures were on average 0.6°C warmer compared with that of N-applied treatments across different tillage and residue removal rates. There were no seasonal trends observed for soil water content (Fig. 4). No-till on average had 0.37 m3 m−3 soil water content throughout the year, with 0.05 m3 m−3 greater than CT. In 2011, a similar seasonal trend in soil temperature occurred as in previous years, although there was a significant tillage by residue removal interaction effect on soil temperature throughout the year in the AC site (Fig. 5). No-till with 50 and 100% residue removal was 1.1°C warmer than with no residue removed under NT and CT across all residue removal rates, which averaged 20.9°C. There Fig. 4. Tillage system and residue removal rate effects on daily soil temperature and water was less rainfall in 2011 in the AC than previcontent at time of greenhouse gas measurement by nitrogen fertilization rate during the growing ous years, resulting in lower soil water content season at the Armstrong southwest site (ASW) in 2010. (Fig. 5). No-till on average had 0.27 m3 m−3 mid-July 2010 (Fig. 3). There was a significant tillage and tillage soil water content throughout the year with by residue removal effect on soil temperature starting as soil be0.04 m3 m−3 greater than CT. In the ASW site for 2011, there gan to thaw until mid-June. Soil temperatures under NT were on were no significant differences between tillage, residue removal, average 1.1°C cooler when compared with CT. When no residue and N fertilization treatments throughout the year on soil temwas removed under NT, soil temperature was on average 2.2°C perature (Fig. 6). There was less rainfall compared with 2010 in cooler than when 50 and 100% of the residue was removed. this site, resulting in lower soil water content (Fig. 6). No-till on Under CT, there were no significant differences in soil temperaaverage had 0.28 m3 m−3 soil water content throughout the year, ture across the different residue removal rates. After late June, with 0.02 m3 m−3 greater than CT on average. when the corn canopy had completely covered the soil surface Seasonal and Cumulative Soil Surface CO2 Emission in N-applied treatments, there were no significant differences in soil temperatures across tillage and residue removal rates for the The effects of residue removal on soil surface CO2 emission rest of the year. There was a significant N fertilization effect on varied by site, year, tillage, and N fertilization rate. In 2009, there soil temperature starting in early July and until late September. were a significant tillage, tillage by residue removal, and N fertilization rate effects on soil surface CO2 emission for the AC site (Fig. 7). Soil surface CO2 emission followed a parabolic trend, peaking in mid-July. Early in the growing season, CT typically had higher soil surface CO2 emission than NT, until mid-July when corn canopy completely covered the soil surface. Afterward, NT was slightly higher or not significantly different from CT. Average soil surface CO2 emission during the entire growing season was higher under CT at 2.50 CO2 kg ha−1 d−1 compared with NT at 2.24 CO2 kg ha−1 d−1 across all N fertilizations and residue removal treatments. Additionally, there was a higher mean soil surface CO2 emission under fertilized treatments (170 and 280 kg N ha−1) compared with when no Fig. 5. Tillage system and residue removal rate effects on daily soil temperature and water content at time N was applied early in the growing seaof greenhouse gas measurement by nitrogen fertilization rate during the growing season at AC in 2011.

www.soils.org/publications/sssaj

617

son. However, starting in early July, there were no significant differences in soil surface CO2 emission across N fertilization rates for the rest of the year. Significant differences in annual cumulative soil surface CO2 emission were only observed across tillage practices where CT (17.4 CO2 Mg ha−1 yr−1) was 14% greater than that with NT (14.9 CO2 Mg ha−1yr−1) (Fig. 8). In 2010, differences in soil surface CO2 emission were mainly due to tillage, residue removal rate, and N fertilization rate main effects in the AC site (Fig. 7). Similar soil surface CO2 emission seasonal trends were observed as in 2009. Treatments under CT (3.12 CO2 kg ha−1 d−1) had higher soil surface CO2 emission on average than Fig. 6. Tillage system and residue removal rate effects on daily soil temperature and water content that with NT (2.85 CO2 kg ha−1 d−1) at time of greenhouse gas measurement by nitrogen fertilization rate during the growing season at ASW in 2011. throughout the growing season. There were no significant differences between 0% Additionally, annual cumulative soil surface CO2 emissions were (3.22 CO2 kg ha−1 d−1) and 50% (3.06 CO2 kg ha−1 d−1) residue greater from CT system (17.9 CO2 Mg ha−1 yr−1) than that removal rate effects on soil surface CO2 emission, although it was from NT system (16.5 CO2 Mg ha−1 yr−1). There were no dif−1 −1 significantly higher than that from 100% (2.68 CO2 kg ha d ) ferences in cumulative soil surface CO2 emission across residue residue removal treatment. There were also no significant differences between 170 (3.13 CO2 kg ha−1 d−1) and 280 kg N ha−1 (3.24 CO2 kg ha−1 d−1) fertilization N rate effects on soil surface CO2 emission; however, when no N was applied, soil surface CO2 emissions were nearly 20% lower throughout the growing season. Annual cumulative soil surface CO2 emissions were greater from CT system (17.8 CO2 Mg ha−1 yr−1) compared with that from NT system (16.3 CO2 Mg ha−1 yr−1) and increased with N fertilization rate increase, but decreased with residue removal rate increase (Fig. 8). In the ASW site (a well-drained soil), similar seasonal soil surface CO2 emission trends were observed as in the AC site (poorly drained soil), although at lower rates and peaks in early July (Fig. 9). Differences in soil surface CO2 emission were mainly due to tillage, residue removal rate, and N fertilization rate main effects. Annual cumulative soil surface CO2 emissions were greater in the ASW (Fig. 10) than in the AC site because of a longer growing season result- Fig. 7. Tillage system and residue removal rate effects on daily soil surface CO2 emissions by ing in warmer soil temperatures (Fig. 4). nitrogen fertilization rate during three growing seasons at AC. T, tillage; P, planting; N, ureaammonium nitrate application; H, harvest; R, residue removal; FT, fall tillage.

618

Soil Science Society of America Journal

removal rates, although CO2 emissions were higher when N was applied than from treatments that did receive N fertilization. In 2011, differences in soil surface CO2 emissions were mainly due to tillage and residue removal rate main effects in the AC site (Fig. 7). Similar soil surface CO2 emission seasonal trends were observed as in previous years. However, CT (3.02 CO2 kg ha−1 d−1) had lower soil surface CO2 emission than NT (3.32 CO2 kg ha−1 d−1) throughout the growing season, which differed from the previous 2 yr. In addition, residue removal rate effects on soil surface CO2 emission also differed from the

previous 2 yr. Greater soil surface CO2 emissions were observed when 50 and 100% of the residue was removed compared with no residue removal treatment. There were also no significant differences between N fertilization rate effects on soil surface CO2 emissions. This resulted in annual cumulative soil surface CO2 emission being higher in NT (18.8 CO2 Mg ha−1 yr−1) compared with CT (17.8 CO2 Mg ha−1 yr−1) (Fig. 8). Additionally, the greatest annual cumulative soil surface CO2 emission occurred when 50% of the residue was removed (19.8 CO2 Mg ha−1 yr−1) followed by 100% removal (18.0 CO2 Mg ha−1 yr−1) and lowest when no residue was removed (17.5 CO2 Mg ha−1 yr−1). However, in the ASW site, there were no significant differences between tillage systems, residue removal rates, and N fertilization rates on soil surface CO2 emission throughout the growing season (Fig. 9). This also resulted in little difference between tillage systems, residue removal rates, and N fertilization rate effects on annual cumulative soil surface CO2 emissions (Fig. 10).

Seasonal and Cumulative Soil Surface N2O Emission

Fig. 8. Annual cumulative soil surface CO2 emissions as affected by tillage system, residue removal rate, and nitrogen fertilization rate at AC from 2009 through 2011. Means with the same letter within each management treatment and year are not significantly different at P £ 0.10. www.soils.org/publications/sssaj

Soil surface N2O emission was only measured at the AC site in 2009 and 2010 (Fig. 11). Differences in tillage systems, residue removal rates, and N fertilization rate effects on soil surface N2O emission were observed after 1 wk of N was applied in both years. As expected, soil surface N2O emission increased as N fertilization rate increased. The largest soil surface N2O emission peak occurred approximately 2 wk after N application, corresponding with the first rainfall event after N application in drier soil conditions for both years. Soil surface N2O emissions throughout the growing season and peaks in general were greater in 2009 than in 2010. Smaller soil surface N2O emission peaks coincided with thawing of the soil and significant rainfall events (Dietzel et al., 2011). In 2009, greater soil surface N2O emission rates were observed from CT compared with NT before the corn canopy had completely covered the soil surface; however, no significant differences were observed thereafter. Additionally, in 2009, although not significantly different because of high variability, removal of 100% residue from the soil surface typically had lower soil surface N2O emission compared with 0 and 50% removal throughout the growing season. In 2010, there were no consistent trends associated with tillage system or residue removal rate effects on soil surface N2O emission. The main contributing factor to higher N2O emission was N fertilization rate in 2009 and 2010 (Fig. 11). As expected, cumulative soil surface N2O emission increased as N fertilization rate and soil mineral N concentration increased in both years. On average, 4.5 and 5.8% of N applied was lost as soil surface N2O emission with 170 and 280 kg N ha−1 fertilization rates, respectively, in both years. In 2009, CT had greater cumulative soil surface N2O emissions compared with NT, with 0 and 50% residue removal in N applied treatments. There were no significant differences in cumulative soil surface N2O emissions between CT and NT with 100% residue removal. In 2010, there were no consistent trends associated with tillage system or residue removal rate effects on cumulative soil surface N2O emission. 619

DISCUSSION

Management Effects on CO2 Emission

Under management practices in CT system with N fertilization rate of 170 kg N ha−1 with no residue removal as in the AC site, annual cumulative CO2 emission contributes, on average, 18.6 CO2 Mg ha−1 yr–1 from 2009 through 2011 (Fig. 8). Under NT, the number is lower: 16.9 CO2 Mg ha−1 yr−1. Under the same management practices in the ASW site from 2010 through 2011, annual cumulative CO2 emission averaged 21.6 CO2 Mg ha−1 yr−1 under CT and 18.9 CO2 Mg ha−1 yr−1 under NT (Fig. 10). Mean annual cumulative CO2 emissions observed in this study are within range Fig. 9. Tillage system and residue removal rate effects on daily soil surface CO2 emissions by nitrogen fertilization rate during two growing seasons at ASW. (12.2 to 25.9 CO2 Mg ha−1 yr−1) from previous studies done in Iowa surface CO2 emissions improved under the nonactive period, but (Parkin and Kaspar, 2006; Hernandez-Ramirez et al., 2009). decreased during the active root activity period. Greater cumulative CO2 emission in CT compared with NT As number of hectares in continuous corn continues to could be attributed to mineralization of soil organic matter and grow because of the increased demand for grain and residue for root derived CO2 respiration, which often follows a parabolic animal feed and ethanol production; changes in fertilization function as affected by soil temperature and water content dursuch as increases in N application are expected (Halvorson et al., ing the growing season (Fortin et al., 1996). In 2009 and 2010, 2006). Annual cumulative CO2 emission was not significantly CO2 emission peaks were more pronounced and higher than in different for 170 N kg N ha−1and 280 N kg ha−1 rates with no 2011 (Fig. 7 and 9). This could be attributed to drier soil condiresidue removal across all tillage systems in both sites. It was extions in 2011 resulting in the broader, shorter peaks of CO2 emispected that annual cumulative CO2 emission would increase besions due to more extreme soil moisture cycling (Mosier et al., cause of greater C input from above- and belowground biomass 2006; Hernandez-Ramirez et al., 2009). Higher soil temperatures due to increasing microbial activity and C mineralization (Dick, under CT compared with NT, during field cultivation to prepare 1992), in addition to increase of root respiration with higher N for planting early in the growing season (Al-Kaisi and Yin, 2005), fertilization rates. Although increases in root respiration were is the primary reason for CT having greater annual cumulative observed with greater N fertilization rate (Guzman and Al-Kaisi, CO2 emissions than NT in 2009 and 2010, even though NT typi2014), differences between 170 and 280 kg N ha−1 fertilization cally has similar or slightly higher soil temperature from July to effects on annual cumulative CO2 emission might have been offOctober (Licht and Al-Kaisi, 2005; Al-Kaisi and Kwaw-Mensah, set by slightly cooler soil temperatures due to shading by corn 2007). Although soil temperature and water content do well to canopy and additional residue left on the soil surface with higher explain differences in management practices effects (Table 1) on a N rates. There is no consensus on how N fertilization effects C daily scale (Linn and Doran, 1984; Davidson et al., 1998), seasonal mineralization and CO2 emission in the field, with some studies trends of soil temperature and soil surface CO2 emissions did not showing a suppressive effect of N on C mineralization and others always coincide with each other. For instance, soil temperatures showing a stimulatory positive effect (Kowalenko et al., 1978; peaked in late June before corn canopy had completely covered Al-Kaisi et al., 2008). the soil surface (Fig. 1, 3, and 5), but CO2 emissions peaked in The effects of residue removal on CO2 emission varied with mid to late July (Fig. 7). Differences between soil temperature and weather conditions. In 2011, when it was considerably drier CO2 emission peaks are attributed to plant growth and root respithroughout the year when compared with 2009 and 2010 (Fig. ration continuing and peaking in late July, which is supported by 1, 3, and 5), the removal of residue increased soil temperature in root respiration measurements done in these study plots (Guzman NT so that it was greater than under CT. Although, soil temand Al-Kaisi, 2014) and as suggested by Sainju et al. (2012) in peratures were not significantly different under NT when no their study. Consequently, when annual soil surface CO2 emission residue was removed from the soil surface compared with CT was divided by active and nonactive root activity periods (Table under all three residue removal rates (Fig. 5). This could explain 1), soil temperature and water content role as predictors for soil why NT with greater than 50% residue removal had larger an620

Soil Science Society of America Journal

fer between residue removal treatments across all tillage and N fertilization rates (Fig. 9 and 10). The dissimilarity between the two sites could be attributed to differences in soil texture and drainage class (Feiziene et al., 2010), where the ASW site is well-drained soil and had small differences in soil temperature and water content across tillage treatments, between residue removal rates, and N fertilization rates compared with the AC site, a poorly drained soil. It is important to keep in mind that these findings are for the first 3 yr since establishment of management practices, and that the long-term removal of residue is expected to reduce soil organic C, which can have significant implications for CO2 emission and overall GHG mitigation strategies (Six et al., 2004, Guzman, 2013).

Management Effects on N2O Emission

Fig. 10. Annual cumulative soil surface CO2 emissions as affected by tillage system, residue removal rate, and nitrogen fertilization rate at ASW in 2010 and 2011. Means with the same letter within each management treatment and year are not significantly different at P £ 0.10.

nual cumulative CO2 emission than CT in 2011 across all N fertilization rates (Fig. 8) in the AC site. In 2009 and 2010, where soil water content was lower and soil temperatures were cooler as compared with 2011, management practices with residue removal had lower CO2 emission than those practices where no residue have been removed during the growing season in the AC site (Fig. 7). However, significant differences between residue removal treatment effects on annual cumulative CO2 emission were not observed, perhaps because of the field spatial variability of soil surface CO2 emission measurements throughout the year. In the ASW site, CO2 emission throughout the growing season and annual cumulative CO2 emission did not significantly difwww.soils.org/publications/sssaj

Nitrogen fertilization had the largest impact on N2O emission. In general, the higher N fertilization rate, the greater N2O emission (Fig. 12). Increased N substrate availability due to N fertilization has been known to increase N2O emission because of enhanced nitrification (Robertson et al., 1989). Mean annual cumulative N2O emissions in central Iowa under N fertilization rates of 170 N kg ha−1 for continuous corn with no residue removal was 5.7 N2O kg ha−1 yr−1 in CT and 5.0 N2O kg ha−1 yr−1 under NT. These results are lower in Iowa than previously cited literature under similar management practices that range from 6.8 to 15.4 N2O kg ha−1 y−1 (Hernandez-Ramirez et al., 2009). These differences may be attributed to differences in gas measurements and calculation methodology, source and placement of N, weather, and soil type (Drury et al., 2006; Parkin and Venterea, 2010). Increases in N fertilization rates from 170 to 280 kg N ha−1 resulted in annual cumulative N2O emission increase by 50% when averaged across residue removal rates and tillage systems (Fig. 12). Nitrous oxide emission (Fig. 11) showed a parabolic trend coinciding with mineral N concentration in the soil (Fig. 2) but did not coincide with CO2 emission or soil temperature seasonal trends. Peak N2O emissions following N fertilization are primarily due to nitrification processes (Mosier et al., 2002). Additionally, studies have shown that most of the N2O emission induced by application of N fertilizer occurs within the period of 15 to 40 d, which was the case for both years in this study (Bremner et al., 1981). However, a high correlation was found between soil surface N2O emission and soil temperature, soil water content, and mineral N throughout the year (Table 2). Intraseasonal differences between management practices can be largely explained by differences in soil temperature and water content conditions in addition to mineral N concentrations in the soil, which generally had a positive effect on N2O emission (Parsons et al., 1991). In 2009, annual cumulative N2O emission was greater in CT systems than those of NT in the AC site (Fig. 12). These findings are different from many studies that have shown NT systems typically have higher N2O emission because of greater denitrification (Payne, 1981) and more frequent under high moisture conditions compared with CT systems (Mosier et 621

al., 2002). However, for this study in 2009, soil water content during the growing season was typically above 0.32 m3 m−3 for both CT and NT systems (Fig. 1). Instead, this can be attributed to greater soil temperature early in the growing season associated with CT treatments (Fig. 1) resulting in higher annual cumulative N2O emissions (Fig. 12). Increases in soil temperature have been shown to stimulate microbial activity and N mineralization resulting in greater N2O emission (Mosier et al., 2002). Similar soil water content conditions occurred in 2010, although mean soil temperature was higher than in 2009, but not significantly different between CT and NT (Fig. 3). Consequently, there were no sig- Fig. 11. Tillage and residue removal effects on daily soil surface N2O emissions by nitrogen fertilization rate during two growing seasons at AC. nificant differences in annual cumulative N2O emissions between CT al. (2000) based on SOC sequestration, agronomic inputs, and and NT in 2010 (Fig. 12). GHG emissions for the AC site only (Table 3). Contribution of Removal of residue was expected to lower annual cumulaCH4 to total GHG emissions was not included because of being tive N2O emission because of decrease in soil moisture content only a minor emitter compared with CO2 and N2O emissions and potential N mineralization from remaining residue (Toma (less than 1 and 10%, respectively). The impact of CO2 emissions and Hatano, 2007). However, there were no significant differfrom management inputs such as fuel usage for tillage, seed bed ences in N2O emission observed across residue removal rates in preparation, planting, herbicide applications, residue removal, this study in both years for the AC site (Fig. 12). Again, this was and the use of CO2 producing subsides such as the manufacture probably due to soil water content conditions typically not beof N fertilizer were evaluated (Table 3). Since units of these difing the limiting factor for N2O emission across residue removal ferent inputs vary (i.e., L, oz, MJ), conversion factors were used rates (Fig. 1 and 3). Differences across residue removal rates are to express in CO2 into equivalent (CO2equ) for comparisons beexpected when moisture conditions and N substrate are limiting tween inputs and outputs (Lal, 2004). factors for production of N2O (Sainju et al., 2012). In general, CT systems under continuous corn with or without residue removal were on average net sources for GWP Management Effects on Global Warming Potential (ranging from 3377 to 1086 CO2eq kg ha−1 yr−1), with the only The GWP under different residue removal management sysexception of CT with no N fertilization and no residue removal tems was estimated following the methodology of Robertson et (−49 CO2eq kg ha−1 yr−1). However, because of the large variTable 1. Soil surface CO2 emissions (kg ha−1 d−1) as predicted by soil temperature (Stemp), soil water content (qv), and their interactions across all years, sites, treatments, and whether roots are active (mid-May to Mid-October) or nonactive (January to May and October to December). Whole Year Term

Active Roots

t(Prob > t)

Estimate

Estimate

Nonactive Roots

t(Prob > t)

Estimate

t(Prob > t)

Stemp

—————————————————— CO2 kg ha−1 —————————————————— 8.07 (