Effect of drip irrigation frequency, nitrogen rate and mulching on nitrous oxide emissions in a semi-arid climate: An assessment across two years in an apple orchard

Agriculture, Ecosystems and Environment 235 (2016) 242–252

Contents lists available at ScienceDirect

Agriculture, Ecosystems and Environment journal homepage: www.elsevier.com/locate/agee

Effect of drip irrigation frequency, nitrogen rate and mulching on nitrous oxide emissions in a semi-arid climate: An assessment across two years in an apple orchard Mesfin M. Fentabila , Craig F. Nichola,* , Melanie D. Jonesb , Gerry H. Neilsenc, Denise Neilsenc , Kirsten D. Hannamb a Department of Earth and Environmental Sciences and Physical Geography, University of British Columbia Okanagan, Kelowna, British Columbia, V1V 1V7, Canada b Department of Biology, University of British Columbia Okanagan, Kelowna, British Columbia, V1V 1V7, Canada c Agriculture and Agri-Food Canada, Summerland Research and Development Centre, Summerland, British Columbia, V0H 1Z0, Canada

A R T I C L E I N F O

Article history: Received 22 November 2015 Received in revised form 5 August 2016 Accepted 28 September 2016 Available online 4 November 2016 Keywords: Nitrous oxide emission Drip irrigation Irrigation frequency Nitrogen amount Mulch Orchard Apple

A B S T R A C T

Micro-irrigation scheduling and mulch application can be used by orchardists to match water supply to plant demand and conserve water. There is little information on how these management practices affect nitrous oxide (N2O) emissions from orchard soils, and most previous studies were short-term (<3 months during the growing season). We investigated (1) N2O emissions across a 2 year cycle of orchard management and (2) seasonal N2O emissions by analysing measurements taken before, during and after the growing season in an apple (Malus domestica Borkh) orchard under various managements in a semiarid climate. Treatments included drip irrigation frequency (every day or every 2nd day) delivering the same total amount of water, orchard floor management (bare soil or shredded bark and wood mulch) and nitrogen application rate applied as calcium nitrate by fertigation (20 or 40 g N tree
1). Over a period of two complete years, irrigation every 2nd day reduced area-scaled N2O emissions by 27% and application of shredded bark and wood mulch reduced area-scaled N2O emissions by 19%, suggesting that reduced drip irrigation frequency and mulching may provide an opportunity for suppressing N2O emissions from drip irrigated orchards. Treatment effects on N2O emissions were variable across seasons and years and a significant portion (17–51%) of the annual N2O emissions occurred during the pregrowing season particularly during freeze-thaw cycles, affirming the importance of year round monitoring when assessing the effect of managements on N2O emissions. ã 2016 Elsevier B.V. All rights reserved.

1. Introduction Agriculture is responsible for 84% of the global anthropogenic N2O and 52% of methane (CH4) emissions (Smith et al., 2008). Crop production is believed to contribute only small net emissions of carbon dioxide (CO2) since crop fields act both as sources and sinks for CO2 (Carlisle et al., 2010 and Smith et al., 2008). There are limited data on N2O emissions from apple orchards (Pang et al., 2009), yet these systems cover more than 25 million ha worldwide (FAO, 2013). In Canada, apple orchards cover 15,000 ha (FAO, 2013),

* Corresponding author. E-mail addresses: mesfi[email protected] (M.M. Fentabil), [email protected] (C.F. Nichol), [email protected] (M.D. Jones), [email protected] (G.H. Neilsen), [email protected] (D. Neilsen), [email protected] (K.D. Hannam). http://dx.doi.org/10.1016/j.agee.2016.09.033 0167-8809/ã 2016 Elsevier B.V. All rights reserved.

with over 3,200 ha located in the semi-arid Okanagan region of British Columbia (Seymour, 2015). To improve water-use efficiency, many apple orchards in the Okanagan have converted from overhead and under-tree sprinklers to under-tree micro-irrigation (micro-sprinkler, micro-spray or drip). This trend is expected to continue as demand for irrigation water increases due to climate change. When using under-tree micro-irrigation, nitrogen (N fertilizer) can be applied through the irrigation system (fertigation) during the period of high tree N demand in order to increase N use efficiency. However, fertigation also increases the localized concentrations of N and water, especially in drip-irrigated systems (Smart et al., 2011), and may increase the potential for N2O emissions (Smart et al., 2011 and Zebarth et al., 2008). The effects of under-tree micro-irrigation and fertigation on N2O emissions are not well known.

M.M. Fentabil et al. / Agriculture, Ecosystems and Environment 235 (2016) 242–252

243

2. Materials and methods

were ‘experimental trees’. A common orchard grass mix was grown in the 1.5 m wide inter-row (alley). The ‘row’ part of each plot was kept weed-free via the use of herbicides (primarily glyphosate). The treatments for this study were established in 2012 as a split-plot experiment; the main plot units had a 2  2 factorial design with two irrigation frequencies and two N application rates. The split-plot units within each main plot consisted of three randomly assigned orchard floor management treatments: herbicide treated bare soil (Clean), shredded bark and wood mulch (Mulch) and black plastic woven geotextile (Geotextile). The whole experiment was a randomised complete block experiment with all treatment combinations replicated in six blocks; only three of the six blocks were used for greenhouse gas study in the present study, with the other three in use by other parallel experiments. Irrigation was designed to deliver 100% of the water lost to evapotranspiration via drippers (4L h
1) located 0.30 m on either side of each tree and suspended approximately 0.3 m above the soil surface. An atmometer (ETGage Co, Loveland, CO) measurement of potential evapotranspiration and a crop coefficient model was used to estimate actual evapotranspiration (Neilsen et al., 2015); the system was controlled automatically by a CR10X datalogger (Campbell Scientific, Logan, UT). The irrigation treatments consisted of irrigating twice a day (morning and afternoon) every day, or irrigating twice a day (morning and afternoon) every second day. The irrigation season extended from May through October and all plots received the same quantity of water. Plots were fertigated with Ca(NO3)2 for six weeks, from around mid-May to the end of June between 2012 and 2014; irrigation and fertigation were timed to coincide with apple development. Treatments were 20 N g tree
1 or 63 kg N ha
1 (Low N, LN) and 40 N g tree
1 or 127 kg N ha
1 (High N, HN). The Clean and Mulch orchard floor management treatments were assessed in this study. The Geotextile treatment was excluded. Mulch was applied on a 2-m wide strip centered on the apple tree row. It was composed of shredded bark and woodchips (primarily from Pinus contorta var latifolia and Picea glauca) generated as waste from local sawmills. It was surface applied in late May of every second year (2012, 2014) to maintain a full mulch with a depth of approximately 10 cm. New mulch was topped up on existing mulch without disturbing the underlying material. Mulch was not applied in 2013 because it was still sufficiently thick to suppress weed growth.

2.1. Study site and experimental design

2.2. Soil sampling and analyses

The study was conducted at the Summerland Research and Development Centre (SRDC) of Agriculture and Agri-Food Canada (Lat. 49 340 N and Long. 119 380 W), located in the Okanagan Valley near Summerland, BC, Canada. The site has a 30-year average (1981–2010) annual precipitation of 346 (25) mm, and daily average temperature of 9.6 (2.4)  C with a minimum daily average of around
2  C in January and a maximum daily average of around 28  C in July. (Environment Canada 2014a; 2014b). The soil is classified as an Osoyoos loamy sand (Wittneben, 1986), which is glacio-fluvial in origin, had a low water-holding capacity, a cation exchange capacity (CEC) of 7.9 meq (100 g)
1, a pH of 6.6, and a C:N ratio of 7.9 before treatment initiation. The soil at the site is typical of apple orchards in the Okanagan Valley. The experiment was conducted in an Ambrosia apple orchard planted in 2003. The apple trees were arranged in eight 53.1 m long rows with a 0.9 m tree spacing within rows and a 3.5 m spacing between rows. The outer row on each side of the planting were ‘guard rows’, used to eliminate edge effects and were not directly involved in the experiment. Each row (block) consisted of twelve 5tree plots with dimensions of 4.50 m  2 m. The trees on each end of the individual plots were ‘guard trees’ and the middle three trees

Soil sampling was conducted in May, June, August, and October in 2013, and every month from April to August, and October in 2014 for a total of 10 soil sampling rounds over two years. At each soil sampling time, soil cores from nine locations to a depth of 15 cm were collected in each of the row and in the alley using a 2-cm diameter auger (Fig. 1). Samples were separately composited for the row and the alley part of each plot. Soil samples were kept frozen until extraction and analysis. Determination of available nitrate-N and nitrite-N (hereafter referred to as NO3
-N because concentrations of NO2
-N were minimal) and ammonium-N (NH4+-N) on soil extracts used 2 M KCl in a 1:5 soil to extractant ratio with a 1-h shaking time, followed by filtration through Whatman No. 40 filter paper. The concentrations of NO3
-N and NH4+-N in the extracts were determined using a segmented flow analyzer (SFA, Model 305D, Astoria Pacific International, Clackamas, OR). Salt-extractable organic C (SEOC) was extracted in the same way as NO3
-N and NH4+-N but a 0.45mm membrane filter (Millipore Corp, USA) was used for filtration (Chantigny et al., 2008), and an Aurora 1030W OI Analytical TOC analyzer (OI Analytical, USA) was used to determine concentration of SEOC.

Orchardists employ various irrigation frequencies, N application rates, organic amendments and orchard floor managements (e.g.: geotextile or mulch ground covers) to increase water and nutrient-use efficiency and improve fruit quality and productivity. These management practices influence soil moisture and nutrients (C and N) dynamics which in turn affect N2O production via nitrification and denitrification. Higher N application rate usually results in higher N2O emissions (Shcherbak et al., 2014). In contrast, varying irrigation frequency and applying organic amendments has been shown to have inconsistent effects on N2O emission. Rolston et al. (1982) found the greatest total N2O emission from perennial ryegrass occurred under the most frequent irrigation while Abalos et al. (2014) found no relationship between irrigation frequency and N2O emissions for melons. Organic matter addition resulted in decreased N2O emission (Livesley et al., 2010; Lopez-Fernández et al., 2007; Nyamadzawo et al., 2014; Sanchez-Martin et al., 2010; Steenwerth and Belina, 2010) while in other studies increased N2O emission (Cochran et al., 1997; Laidlaw, 1993; Pelster et al., 2012). Most of these prior studies either measured N2O emissions over a short period during the growing season or used manure as the source of organic matter (Fentabil et al., 2016). A 2 year study of the use of shredded bark and wood mulch applied as ground cover in a grape (Vitis vinifera L.) vineyard reduced cumulative N2O fluxes by 28% (Fentabil, 2016; Fentabil et al., 2016); however, further work is needed to substantiate these results over a wider range of experimental conditions and crops. We assessed the impacts of two frequencies of drip irrigation, two N application rates and two orchard floor managements (bare soil or shredded bark and wood mulch) on N2O emissions and soil characteristics in an apple orchard. We hypothesized that higher irrigation frequency and higher N application rate would result in greater N2O emissions due to increased soil moisture and N while application of a shredded bark and wood mulch as a soil cover would result in a decrease in N2O emissions due to higher microbial immobilization of mineral N in the presence of a labile carbon source. Our objectives were to assess (1) seasonal N2O emissions before, during and after the growing season and (2) cumulative N2O emissions over a 2 year cycle of orchard management.

244

Alley = 0.75m

M.M. Fentabil et al. / Agriculture, Ecosystems and Environment 235 (2016) 242–252

4L/hr dripper Approximate Wetted area

Row = 1.00m

Apple tree

Alley = 0.75m

Row = 1.00m

Apple tree row

Apple tree row x

x

x

x

x

x

x

x

x

Alley chambe r

0.45m

Two adjacent 2L/hr dripper Row Chamber

y

y

y

y

y

y

y

y

y

Campbell CS616 moisture content sensor and type-T thermocouples

0.45m

Fig. 1. Spatial representation of a single apple tree functional unit in a drip irrigated plot. “x” and “y” represent examples of soil sampling locations for row and alley for one sampling event, respectively. Similar locations in other experimental trees (not shown) in the same plot were sampled for each soil sampling event in May, June, August, and October in 2013 and every month from April to August, and October in 2014.

Total C and N content were measured by combusting 15–20 mg finely ground air dried sample using a Costech 4010 Elemental Analyzer with thermal conductivity detection. CEC was measured using an ammonium acetate extract buffered to pH 7. Soil pH and EC were measured on extracts of 1:2 (wv
1) deionized water: finely ground, sieved (2 mm) and air-dried soil using a pH meter (WTW inoLab pH 7200) and an EC meter (WTW inoLab Cond 7200). In-situ volumetric soil water content (0–30 cm depth) and temperature (at 2 cm, 10 cm, 20 cm and 50 cm depth) for each plot were measured continuously at 1-h intervals using 30 cm length time domain reflectometry (TDR) probes installed vertically (Campbell Scientific, CS616) and type-T thermocouples (Omega Engineering, Stamford, CT), respectively, monitored using a Campbell Scientific CR1000 data logger. Thermocouples were positioned directly adjacent to the TDR probes at 2 cm, 10 cm, 20 cm and 50 cm depths in each of the plots. All TDR probes and thermocouples were installed permanently at the center of each plot where N2O flux was measured, and kept undisturbed throughout the experiment. Only the 2 cm thermocouple data are discussed further because N2O emissions are known to be mainly affected by the temperature of the top 5 cm of soil. Soil bulk density samples were collected at two locations, 15 cm and 30 cm away from the dripper parallel to the tree row to represent the fertilized strip, and at the center of the alley. All soil bulk densities were sampled from 3 to 9 cm depth using 6 cm diameter copper collars fitted to a soil bulk density sampler. An estimate of the fraction of water filled pore space (WFPS) was calculated as described in Fentabil et al. (2016). 2.3. N2O flux measurements Gas flux monitoring started in January 16, 2013 and ended on December 22, 2014. The regular monitoring schedule involved

sampling: twice a week during fertigation; once a week during irrigation in the fall and in the spring; and once every second week in winter. Sampling frequency was increased to two or three times a week during short-term weather changes such as spring thaw or intense rainfall and during management events such as mulch application, irrigation initiation, and fertigation initiation. N2O gas samples were collected using non-flow-through non-steady-state (NFT-NSS) chambers (Rochette and Eriksen-Hamel, 2007). Rectangular aluminum frames (0.69 m  0.40 m  0.15 m depth) were installed to a depth of 0.13 m one month prior to the beginning of monitoring and left undisturbed for the duration of the experiment (Fig. 1). The 4 L h
1 dripper at the chamber edge was replaced by two 2 L h
1 drippers, one dripping on either side of the edge of the chamber. Gas samples (20 mL gas in 12 mL pre-evacuated Exetainers, Part No: 737W, Labco Ltd.) were collected at 0, 7, 14, and 21 min in the summer and at 0, 10, 20, and 30 min in the winter. N2O concentrations were analysed within one week of sampling using a gas chromatograph (Bruker 456, now Scion Instruments, Freemont, U.S.A.) equipped with an electron capture detector (ECD) and a CTC Combi-Pal auto sampler (CTC Analytics AG, Zwingen, Switzerland). The soil-surface N2O fluxes were determined by calculating slope (dC/dt) of concentration (C) vs time (t) at chamber closure (t = 0) via either linear or non-linear regression (as appropriate) using the equations of Rochette and Hutchinson (2005). Non-linear regression (Hutchinson and Livingston, 1993) was used when the accumulation of N2O decreased with time. Linear regression was used when the accumulation of N2O was consistent with time (Rochette and Eriksen-Hamel, 2007). Relative humidity, air temperature, and pressure for flux calculations were obtained from a weather station located within 0.5 km (Environment Canada, 2014b) at the SDRC. Orchard-scaled emissions factors (EFs), uncorrected for background emission for each treatment, were expressed as the percentage of the applied N emitted as N2O-N using: P (1) EF = ( N2Oweighted/Applied available Nsource)  100% P Where: N2Oweighted is the area-weighted annual cumulative N2O emission (kg N2O-N ha
1) calculated by considering the percentage area of the field accounted by the row (57%) and alley (43%) and Applied available Nsource is the application rate of N fertilizer (63 or 127 kg N ha
1). 2.4. Yield-scaled N2O emissions Apples were harvested in late September of each year. Yield (kg ha
1) was determined from the harvest at the three central trees in every plot. Yield-scaled N2O emissions (g Mg
1) for each treatment were determined by dividing the annual cumulative N2O emissions (g ha
1 year
1) by yield (Mg ha
1 year
1). 2.5. Data and statistical analysis Data were analysed to determine fluxes during (i) the spring thaw and the pre-growing season
PreGS (January through April); (ii) the growing season—GS (May through October); (iii) the period when ground freezes (post-harvest or post-growing season— PostGS [November and December]); and (iv) annually (January P through December). Area-scaled N2O for individual plots were calculated using linear interpolation of flux rates between sampling days (Millar et al., 2012) and by extrapolation to per hectare of the fertilized strip. Prior to analysis, all data were corrected for both lack of normality and homogeneity of the error variance by using Box-Cox transformations (Box and Cox, 1964) via the SAS transgress procedure, but all data are reported as untransformed means. Effects of treatments on N2O emissions

M.M. Fentabil et al. / Agriculture, Ecosystems and Environment 235 (2016) 242–252

and environmental variables were determined using the PROC MIXED procedure in SAS, with repeated measurements in the model option (Version 9.3; SAS1 Institute, Inc., Cary, NC); block, and block-by-irrigation-by-nitrogen rate were treated as random effects. Model selection was conducted as described by Littell et al. (2006). Since N2O emissions were measured over unequally spaced time intervals, a spatial power law (SP(POW)) covariance structure was used. The SP(POW) structure for unequally spaced longitudinal measurements provides a direct generalization of the autoregressive model (order one) for equally spaced measurements. The SP(POW) models the covariance between two measurements at times t1 and t2 as: h i Cov Y t1 ; Y t2 ¼ s 2 rjt1
t2 j ð2Þ Where Y is a measurement, r is an autoregressive parameter assumed to satisfy |r| < 1 and s2 is an overall variance. P Similarly, the effects of treatments on N2O were conducted using the PROC MIXED procedure, for individual years, and for the 2-year combined data with years as repeated measures. Compound symmetry covariance structure was used in the “type” statements of the repeated model to draw overall statistical comparisons

245

among management factors imposed over two years. KenwardRoger’s adjustment was used to estimate degrees of freedom (Kenward and Roger, 1997). Pairwise means comparisons were performed using the PDIFF statement and Tukey-Kramer adjustment method. Unless otherwise mentioned, a p-value < 0.05 was used for fixed effects and means separation. 3. Results 3.1. Weather, soil temperature and WFPS Environmental conditions in 2013 and 2014 were typical of the region. The mean daily air temperature in 2014 (9.8  C) was slightly higher than 2013 (9.6  C), which was the same as the 30-year average (9.6  C) (Table 1). In both years, daily mean soil and air temperatures were mostly between 20  C and 30  C in July and August and below 0  C from December through February (Fig. 2). More precipitation fell through the year in 2013 than in 2014, and a greater proportion of the precipitation fell during the fertigation period. The largest individual rainfall events also occurred in 2013 (June 20, 26 mm; June 24, 32 mm). In 2014, a single event of 35 mm was noted on June 13 (Fig. 2). Nevertheless, the total quantity of

Table 1 Summary of climate and soil data during experimental years 2013 and 2014 including thaw dates, precipitation, irrigation, and average air and soil temperatures. Data presented during the pre-growing season (PreGS, January through April), the growing season (GS, May through October), and post-growing season (PostGS, November and December). 2013

Parameter

Date of first thaw Date of last thaw Total precipitation (mm)z Total irrigation (mm) Total water inputsy Average soil temperature ( C)x Average air temperature ( C) No of days soil temperature < 0  C No of days air temperature < 0  C No of days soil temperature <
5  C No of days air temperature <
5  C z y x

2014

PreGS

GS

PostGS

PreGS

GS

PostGS

January 9th Febuary 20th 38/55 – 93 3.1 3.5 36 29 7 6

– – 137/104 145/601 987 17.6 17.2 0 0 0 0

– – 36 – 36 0.6
1.1 36 37 1 12

January 10th March 4th 22/43 – 65 3.7 2.6 40 40 0 14

– – 74/116 148/538 875 18.6 17.5 0 0 0 0

– – 71 – 71 1.6 0.8 20 23 2 9

Data separated by “/” indicates either thaw period/remainder of PreGS totals or fertigation period/ irrigation period totals. Total water inputs = Total precipitation + Total irrigation. Soil temperature was measured at 2 cm depth.

45

I Ferti g..g.

35

Irrigation (I)

Total Precip. (mm)

I Ferti g..g.

Irrigation (I)

Irrigation (mm) Mean Air Temp (°C)

25

Mean Soil Temp (°C)

15 5 -5 -15 1-Jan-13 3-Mar-13 3-May-13 3-Jul-13 2-Sep-13 2-Nov-13 2-Jan-14 4-Mar-14 4-May-14 4-Jul-14 3-Sep-14 3-Nov-14 2013 Pre-Growing Season

2013 Growing Season

2013 Postgrowing Season

2014 Pre-growing Season

2014 Growing Season

2014 Postgrowing Season

Fig. 2. Air and soil temperature, precipitation and irrigation inputs for 2013 and 2014. “Fertig” in the top insert represents periods of fertigation, the application of N through the irrigation system. The two different sized columns (blue) for irrigation are caused by the two irrigation frequencies; the shorter column indicates the total mm of water applied as the result of daily irrigation and the taller column indicates the total mm of water applied as the result of irrigation every day and every 2nd day (applied on even Julian days). Both strategies supply the same total volume, matched to estimated ET. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

246

M.M. Fentabil et al. / Agriculture, Ecosystems and Environment 235 (2016) 242–252

A

Thaw

I Ferg. Irrigaon (I)

Thaw

I Ferg.

75

WFPS Every 2nd day irrig.

65 WFPS (%)

Irrigaon (I)

. WFPS Every day irrig.

55 45 35 25 15

WFPS (%)

B

75

WFPS 40g per tree

65

WFPS 20g per tree

55 45 35 25 15

C

75

WFPS Clean WFPS Mulch

WFPS (%)

65 55 45 35 25 15 1-Jan

3-Mar

2013 Pre-growing season

3-May

3-Jul

2-Sep

2013 Growing season

2-Nov

2-Jan

2013 Postgrowing season

4-Mar

4-May

2014 Pre-growing season

4-Jul

3-Sep

2014 Growing season

3-Nov 2014 Postgrowing season

Fig. 3. Water filled pore space (WFPS) of the soil in 2013 and 2014 across: (A) irrigation frequency (B) N application rate (C) orchard floor management. Vertical dashed lines are used to separate seasons. “Fertig” in the top insert represent periods of fertigation, the application of N through the irrigation system.

irrigation water applied each year was similar. The soil WFPS was similar among treatments (Fig. 3). The WFPS during the 2-year GS averaged 48% across treatments and none of the treatments affected WFPS significantly (p > 0.05). Despite an overall lack of statistical differences among treatments, irrigation frequency had a stronger effect than orchard floor management on the time soil was at a high WFPS. On average during the 2-year GS, the LF plots exceeded 60% WFPS only 4% of the time while the HF plots exceeded 60% WFPS 26% of the time. Our water content sensors were deployed to determine average water content over the top 30 cm for multiple experimental goals; WFPS in the top 5 or 10 cm of the plots, or directly under the drippers and beneath mulch, may have been higher than reported. 3.2. Daily N2O emissions The temporal patterns of N2O emissions (Fig. 4) for the years 2013 and 2014 were characterized by high N2O emissions within two periods: during freeze-thaw cycles in the PreGS and during the irrigation or fertigation period in the GS of each year. Emissions during the PostGS were negligible. For each year, the highest emissions were recorded at the start of irrigation (mid-May 2014), i.e., a week prior to the start of fertigation or following either a

week of intense rainfall (late June 2013) during the fertigation period. Treatment effects on daily N2O emissions were not consistent across seasons within a year and between years (Table 2). Sampling date (Date) consistently accounted for most of the variation in N2O emissions across seasons within a year and between years. During the PreGS in 2013, Mulching significantly reduced N2O emissions on 4 of the 11 monitoring days (Fig. 4C). During the GS in 2013, less frequent irrigation reduced N2O emissions on 5 of the 31 monitoring days (Fig. 4A) while mulching reduced N2O emissions early in summer and increased them later in summer. (Fig. 4C). During the PreGS in 2014, Mulching had more effect on N2O emissions than Irrigation and N-rate; Mulching reduced N2O emissions on 4 of the 13 monitoring days (Fig. 4). During the GS, both LF irrigation and mulching caused significantly lower N2O emissions on 2–6 of the 29 monitoring days (Fig. 4A and C). 3.3. Seasonal and annual cumulative N2O emissions P Treatment effects on cumulative N2O emissions ( N2O) were variable across seasons and years (Table 3). In 2013, LF irrigation P decreased the annual N2O emissions while Mulching reduced P P N2O emissions only during the PreGS. In 2014, the annual N2O

M.M. Fentabil et al. / Agriculture, Ecosystems and Environment 235 (2016) 242–252

247

Fig. 4. Mean daily of N2O emissions in 2013 and 2014 by: (A) irrigation frequency and alley position (B) N application rate (C) orchard floor management. Vertical dashed lines are used to separate seasons. Capped bars indicate standard error of the mean (n = 12). *, **, ***, **** indicate differences of least squares means using the Tukey-Kramer adjustment, at p < 0.05, p < 0.01, p < 0.001, p < 0.0001 respectively. “Fertig” in the top insert represent periods of fertigation, the application of N through the irrigation system. Irrigation and thaw periods also indicated.

Table 2 F-value of repeated measures analyses of daily N2O emissions in an apple orchard during the pre-growing season (PreGS, January through April) and during the growing season (GS, May through October) in 2013 and 2014. Effect

2013

2014 GS

PreGS

Irrigation (I)y N-rate (N) Floor mgmt (F)x Date (D) ID ND FD z y

x

PreGS

GS

F Valuez

Num DF

F Value

Num DF

F Value

Num DF

F Value

Num DF

1.44 0.45 34.51**** 9.06**** 0.76 2.07* 5.53****

1 1 1 10 10 10 10

9.49* 0.18 0.90 28.09**** 1.15 1.34 2.82****

1 1 1 30 30 30 30

2.64 3.49k 3.37k 71.22**** 2.21* 2.28* 2.79**

1 1 1 12 12 12 12

2.70 1.28 19.48**** 10.08**** 1.06 0.80 0.46*

1 1 1 28 28 28 28

k, *, **, ***,**** indicates that differences between treatments are significant at p < 0.1, 0.05, 0.01, 0.001, 0.0001, respectively. Irrigation occurred from May 8 to October 23 in 2013 and from May 12 to October 23 in 2014. Ca(NO3)2 was applied via fertigation from May 14 to June 25 in 2013 and May 21 to July 2 in 2014. Bark and wood mulch was surface applied in May of 2012 and 2014.

248

M.M. Fentabil et al. / Agriculture, Ecosystems and Environment 235 (2016) 242–252

Table 3 Effects of irrigation frequency, N application rate and orchard floor management on seasonal and annual N2O emissions in an apple orchard during the pre-growing season (PreGS, January through April) and during the growing season (GS, May through October) in 2013 and 2014. Effect

Cumulative N2O emission (kg ha
1)z 2013

2014

PreGS

GS

Annual

PreGS

GS

Annual

Irrigation Low Frequency (LF) High Frequency (HF)

0.15  0.04 0.18  0.04

0.63  0.05 0.93  0.12

0.78  0.09a 1.12  0.14b

0.38  0.04 0.51  0.07

0.34  0.04 0.43  0.04

0.76  0.07a 0.97  0.106b

N-rate Low N (LN) High N (HN)

0.17  0.04 0.16  0.04

0.78  0.08 0.78  0.13

0.96  0.10 0.94  0.15

0.48  0.03 0.42  0.08

0.37  0.04 0.40  0.04

0.88  0.05 0.85  0.12

Floor mgmt Mulch Clean

0.08  0.03a 0.25  0.04b

0.80  0.13 0.76  0.08

0.89  0.15 1.02  0.10

0.40  0.04 0.50  0.07

0.32  0.04a 0.44  0.04b

0.74  0.07a 0.99  0.09b

Pr > Fy Irrigation (I) N-rate (N) Floor mgmt (F) IN IF NF INF

ns ns ** ns ns ns ns

k ns ns ns k ns ns

* ns k ns k ns ns

k k ns ns ns ns ns

k ns * ns ns k ns

* ns ** * ns * ns

z

y

Mean N2O emission calculations for treatments were based on a 2 m-wide fertilized strip. Means followed treatments by different lowercase letters within columns indicate differences of least squares means between pairs of using the Tukey-Kramer adjustment, at p< 0.05. k, *, **, ***, **** and ns indicate a significant treatment effect at p 0.10, 0.05, 0.01, 0.001, 0.0001 or no significant effect, respectively.

emissions were reduced by LF irrigation and Mulching. Mulching P also decreased N2O emissions during the GS. Treatment interactions were seldom significant in either year. P There was no difference in annual N2O between years. However, the relative contribution of seasons towards the total P annual N2O varied across years (Table 3). In 2013, the majority P (82%) of the annual N2O emissions occurred during the GS, followed by PreGS (17%); the emissions during the PostGS were negligible (1%). In 2014, both the PreGS (51%) and GS (45%) contributed equally while the PostGS only accounted for 4% of the total N2O emission (Fig. 4 and Table 3). The average N2O emissions

in all the treatments in the PreGS of 2014 (0.45 kg ha
1) were higher than in 2013 (0.17 kg ha
1) while the emissions in the GS of 2013 (0.78 kg ha
1) were much higher than in 2014 (0.39 kg ha
1). Based on combined 2-year data, LF irrigation and mulching P reduced annual N2O emissions by 27% and 19%, respectively, while no difference was found between nitrogen application rates (Table 4). The significance of mulching for reducing N2O emission was also apparent on the combined PreGS and combined PostGS data. Similarly, LF irrigation reduced the N2O emissions during GS and the annual N2O emissions.

Table 4 Effects of irrigation frequency, N application rate and orchard floor management on 2-year mean seasonal and annual N2O emissions, 2-year mean N2O emissions factor (EF, N2O-N emissions per unit of total N applied), and 2-year mean yield-scaled N2O emissions (N2O yield
1) in an apple orchard. Year was included as a repeated measure factor in the analysis using Proc Mixed in SAS. Cumulative N2O emissions (kg ha
1 [season or year]
1)z

PreGS

GS

PostGS

Annual

N2Oy emission factor (%)

Irrigation Low Frequency (LF) High Frequency (HF)

0.27  0.04 0.35  0.05

0.49  0.04a 0.68  0.08b

0.02  0.01 0.02  0.01

0.77  0.05a 1.05  0.08b

0.55  0.05 0.64 0.05

21.7  2.0 30.2  3.0

N-rate Low N (LN) High N (HN)

0.33  0.04 0.29  0.05

0.58  0.06 0.59  0.08

0.02  0.01 0.02  0.01

0.92  0.05 0.90  0.09

0.80 0.03a 0. 39 0.03b

25.5  2.0 26.3  3.3

Floor mgmt Mulch Clean

0.24  0.04a 0.37  0.05b

0.56  0.08 0.60  0.05

0.01  0.00a 0.03  0.01b

0.81  0.08a 1.00  0.07b

0.56 0.06 0.63  0.05

24.6 3.2 27.2  2.2

Pr > Fx Irrigation N-rate Floor mgmt Year

ns ns ** ****

* ns ns ****

ns ns * ****

* ns * ns

k *** ns ns

ns ns ns ***

Effect

z y

x

Yieldscaled N2O (g Mg
1)

Mean N2O emission calculations for treatments were based on a 2 m-wide fertilized strip. N2O emission factor calculations for treatments considered both the fertilized strip and the alley. Pairs of means followed by different lowercase letters within columns indicate differences of least squares means using the Tukey-Kramer adjustment, at p < 0.05. *, **, ***, **** and ns indicate a significant treatment effect at p 0.10, 0.05, 0.01, 0.001, 0.0001 or no significant effect.

M.M. Fentabil et al. / Agriculture, Ecosystems and Environment 235 (2016) 242–252

249

Emission factors (EFs) are N2O emissions per unit of total N applied uncorrected for background emissions. EFs averaged 0.60% across the entire study period and ranged from 0.39% to 0.80% across treatments (Table 4). There were no significant differences caused by irrigation treatment or mulching. Emission factors were affected by N application rates; the EF in HN applications was approximately half that of the EF in the LN application rate. YieldP scaled N2O emissions (ratios of N2O emission to annual yield) were not affected by the treatments imposed. The N2O emissions during the same season but in different years and yield-scaled N2O emissions across years were significantly different (p < 0.01, Table 4).

In both years, average soil SEOC concentrations were not affected by irrigation frequency or N-application rate (Table 5). Mulching increased soil SEOC concentrations by 38% and 99% in 2013 and 2014, respectively. On average, soil SEOC concentrations doubled from 2013 to 2014. None of the management factors affected soil pH across years. The soil became slightly more acidic between 2012 and 2014 (p < 0.0001); soil pH was 6.6 in 2012 before treatment initiation and averaged 6.5 and 6.3 across treatments, in 2013 and 2014, respectively. In 2013, soil EC was not affected by treatment. However, in 2014 mulching reduced soil EC by 22%. Fresh mulch was applied only in 2012 and 2014. On average, soil EC doubled from 2013 to 2014.

3.4. Soil nutrients and chemistry

4. Discussion

The majority (68%) of mineral N was in the form of nitrate. In 2013, average annual soil NO3
-N concentrations in Clean plots were approximately twice those of Mulch plots. Similarly, average annual soil NO3
-N concentrations in HN plots were approximately two times higher than LN plots. (Table 5). In 2014, average annual soil NO3
-N concentrations in Clean plots were five-times higher than Mulch plots. Average annual soil NO3
-N concentrations were not affected by the rate of N application. On average, mean soil NO3
-N concentrations remained similar between years. Average soil NH4+-N concentrations were similar across treatments in 2013 (Table 5), but were higher in Mulch plots in 2014. On average, soil NH4+-N concentrations doubled from 2013 to 2014.

4.1. Effect of irrigation frequency, N rate and mulching on N2O emissions Averaged over two years, our data shows that less frequent irrigation (every 2nd day) reduced N2O emissions by 27% compared to the more frequent irrigation (every day). This difference may have been caused by recurrent periods of high WFPS in the more frequently irrigated plots. Soils maintained at higher volumetric moisture content over a longer period of time may result in higher N2O fluxes than soils subjected to wetting and drying cycles (Rolston et al., 1982). Denitrification is the main direct source of N2O from nitrate-based fertilizers (Russow et al., 2008) and higher

Table 5 Effects of irrigation frequency, N application rate and orchard floor management on mean extractable NO3
-N, NH4+- N, salt-extractable organic carbon (SEOC), pH and electrical conductivity (EC) of soil (0–15 cm depth) from apple orchard during the growing season (May through October) in 2013 (n = 96, i.e., 24 plots  4 sampling events) and April through October in 2014 (n = 144, i.e., 24 plots  6 sampling events). NO3
-Nz (mg kg
1)

NH4+-N (mg kg
1)

SEOC (mg kg
1)

pH

EC (mS cm
1)

Irrigation Low Frequency (LF) High Frequency (HF)

13.7  3.1 11.5  1.6

3.7  0.6 3.3  0.3

244  26 211  18

6.56  0.04 6.44  0.04

100  3 102  5

N-rate Low N (LN) High N (HN)

9.3  1.0a 16.0  3.3b

3.2  0.3 3.8  0.6

230  25 226  21

6.59  0.04 6.40  0.03

103  5 99  4

Floor mgmt Mulch Clean

7.1  0.8a 18.1  3.2b

3.2  0.2 3.8  0.6

264  26a 192  17b

6.51  0.04 6.49  0.04

91  1 111  5

Pr > Fy Irrigation N-rate Floor mgmt

ns ** ****

ns ns ns

ns ns **

ns ns ns

ns ns ns

Irrigation Low Frequency (LF) High Frequency (HF)

10.2  1.6 16.1  3.5

8.3  0.6 8.9  0.8

309  34 362  40

6.33  0.04 6.26 0.04

208  21 216  20

N-rate Low N (LN) High N (HN)

12.5  2.5 13.7  3

8.1  0.6 9.2  0.8

321  37 350  37

6.31 0.04 6.28 0.04

213  19 211  21

Floor mgmt Mulch Clean

4.3  0.4a 21.9  3.5b

9.8  0.9a 7.5  0.4b

446  41a 224  27b

6.29 0.03 6.30 0.04

186  19a 238  21b

Pr > Fy Irrigation N-rate Floor mgmt

ns ns ****

ns ns ****

ns ns ****

ns ns ns

ns ns ***

Year

Effect

2013

2014

z y

Pairs of means followed by different lowercase letters within columns indicate differences of least squares means using the Tukey-Kramer adjustment, at p < 0.05. Within columns *, **, ***, **** and ns indicate a significant treatment effect at p 0.05, 0.01, 0.001, 0.0001 or no significant effect, respectively.

250

M.M. Fentabil et al. / Agriculture, Ecosystems and Environment 235 (2016) 242–252

WFPS over longer periods can promote greater rates of denitrification. Higher irrigation frequency usually results in a smaller wetted soil volume and higher mean soil water content (Wang et al., 2006). Our observation of increased N2O under high frequency irrigation is consistent with the findings of Rolston et al. (1982), who measured N2O emissions in a spray-irrigated perennial ryegrass field, but it differs from the findings of Abalos et al. (2014), who found no effect of irrigation frequency on N2O emissions in a drip-irrigated melon field. However, both studies were for less than 3 months during the growing season; the longerterm effects of irrigation frequency on N2O emissions in those studies are unknown. Our 2-year study indicated that irrigation frequency affects cumulative emissions of the growing season as well as total annual emission. Contrary to our expectations, N application rate did not have a P significant effect on mean 2-year N2O emissions. Although higher nitrogen application rates usually result in higher N2O emissions, the rates of N applied in our orchard study (63 kg N ha
1 or 127 kg N ha
1) were much lower than those applied in other orchard studies with N rates of: 210 kg N ha
1 in Lin et al. (2012); 312 kg N ha
1 in Pang et al. (2009); and 579–597 kg N ha
1 in Lin et al. (2010). This may indicate that fertilizer N usage by the apple trees under fertigation was more efficient leaving less fertilizer N susceptible to emissions as nitrous oxide. No differences were found in leaf or fruit nitrogen content (Hannam, personal communications) but a stable isotope labelled nitrogen study would be needed to determine the details of the source of N uptake by the apple trees. A second factor may be that the applied calcium nitrate fertilizer was transported to lower depths in the soil profile where complete denitrification to N2 was more likely. We note that the soil NO3
concentrations (0–15 cm) in the orchard were approximately three-times lower than those in a nearby vineyard on a slightly finer soil (sandy loam) receiving 40 kg N ha
1 year
1 as urea (Fentabil et al., 2016). The soil temperature in the current study was also much lower than other studies (e.g.: Lin et al., 2010, 2012). The lower soil temperature in combination with lower NO3
P concentrations may have caused the 2-year mean N2O emissions to remain similar irrespective of N application rate. This is consistent with other studies that related N2O production to soil temperature (Benoit et al., 2015) and soil N availability (Van Groenigen et al., 2004). Surface application of shredded bark and wood mulch reduced N2O emissions by 19%. Similar results were observed at a microirrigated vineyard located within 0.5 km of the current research site; in that study, mulching reduced N2O emission by 28% (Fentabil et al., 2016). Mulching also decreased soil NO3
-N concentrations (by 61% in 2013 and by 80% in 2014) and increased soil SEOC in both years at that site. This is consistent with other studies (Fentabil et al., 2016; Hannam et al., 2015; Homyak et al., 2008), where lower soil nitrate was detected under mulch, likely due to increased microbial immobilization of mineral N in the presence of increased available carbon. In the current study, the lower soil NO3
-N levels under mulch likely caused lower N2O emissions. 4.2. Seasonal N2O emissions: thaw and the pre-growing season The temporal patterns of N2O emissions in the two monitoring years were characterized by high N2O emissions during freezethaw cycles in the PreGS and during irrigation and fertigation in the GS. A significant portion (17% in 2013 and 51% in 2014) of the total N2O emissions occurred during the PreGS, particularly during the thaw period. During freeze-thaw events in the PreGS, it is likely anaerobic conditions were created. The pore space filled by water and ice in the PreGS was likely higher than reported in the WFPS measurements using dielectric sensor methods, which do not

register frozen water content. WFPS in the PreGS appears low (<45%) during freeze-thaw cycles but the actual WFPS was likely higher in the top 0–5 cm of the soil because the majority of the 30 cm long TDR probe was located within frozen soil below 5 cm. Soil N availability was not measured during the PreGS; however, during this period plant growth was limited and, thus, tree N uptake was probably limited. This may have led to increased soil N concentrations which, together with elevated soil water content at the soil surface, could have created suitable conditions for enhanced denitrification that lead to N2O spikes in both years. The winter of 2014 was colder than that of 2013, and N2O emissions in the pre-GS were almost three-fold greater in 2014 than in 2013. The lower air temperatures that occurred prior to thaw in 2014 may have resulted in greater cell lysis during winter freeze thereby providing fresh substrates for surviving microbes (Herrmann and Witter, 2002), leading to enhanced N2O emissions during thaw. The 2014 thaw period was also longer in duration and included multiple freeze-thaw cycles, which may have promoted additional cell lysis during the thaw period. Lower freezing temperatures have the potential to kill more soil microbes and damage roots, providing more easily degradable substrates for the surviving microbes during the following thaw event (Koponen and Martikainen, 2004; Neilsen et al., 2001). Together, this may explain why there was a significant increase in N2O emissions during the PreGS of 2014 compared with 2013. 4.3. Seasonal N2O emissions: growing season During the GS, the highest daily fluxes (18–31 g ha
1 day
1) occurred in 2013 following intense rainfall events during fertigation, while the highest daily fluxes in 2014 (6–10 g ha
1 day
1) occurred following the start of irrigation but prior to fertigation. The high fluxes in 2013 probably exceeded those observed in 2014 because the intense rain events in late June 2013 coincided with the end of fertigation, when soil nitrate was at a maximum and, therefore, conditions for denitrification were favourable. No similar spikes in N2O emissions were noted following the single intense rain event that occurred on June 13, 2014, well before the end of the fertigation period. The highest daily fluxes in 2014 occurred following the start of irrigation, probably due to increased mineralization of C and N following rewetting of dry soil (Beare et al., 2009). 4.4. Yield-scaled N2O emission Effective management practices combine agricultural productivity with environmental sustainability. Expressing N2O emissions per unit of yield accounts for both productivity and environmental sustainability and may provide a useful metric for greenhouse gas inventories (Venterea et al., 2011). In our study, area-scaled N2O emissions were significantly decreased by both lower irrigation frequency (irrigation every 2nd day) and surface application of bark and wood mulch. However, none of the treatments had an effect on the 2-year mean yield-scaled N2O emissions. Nevertheless, the lack of statistical significance and actual values of yieldscaled emissions reported here may not be representative. Yieldscaled emissions calculations are dependent on both the variability in N2O emissions (n = 3) and the variability in yield (n = 3). While three replicates had sufficient statistical power to compare treatment effects on area-scaled N2O emissions, a reliable value for yield-scaled emissions may not be achievable with such low statistical power. Nevertheless, this study represents the first report of yield-scaled N2O emissions for an apple orchard. The N2O emissions during the same season, but in different years, as well as yield-scaled N2O emissions across years, were significantly different, implying the importance of multi-year continuous

M.M. Fentabil et al. / Agriculture, Ecosystems and Environment 235 (2016) 242–252

monitoring for obtaining representative N2O emissions and yield data. 5. Conclusion Based on a 2 year study, we conclude that N2O emissions from drip-irrigated apple orchards can be reduced by reducing dripirrigation frequency and applying shredded bark and wood mulch to the row portion of the orchard floor. Nitrogen application rate did not affect N2O emissions, likely because of efficient N uptake, and nitrate movement to lower in the soil profile. A significant portion (17–48%) of the annual emission occurred during the pregrowing season and this season must be taken into account when assessing the effect of managements in climates where freezethaw cycles occur. Acknowledgements This work was funded by the Agricultural Greenhouse Gases Program of Agriculture and Agri-Food Canada (AGGP-AAFC), the Canada Foundation for Innovation (CFI), and the work-study program of University of British Columbia (WS). The authors gratefully acknowledge the contributions of: Dr. Marina Molodovskaya, Valerie Ward, Scott Fazackerley and Markandu Anputhas at UBC; Dr. Tom Forge, Shawn Kuchta, Istvan Losso, Bill Rabie and student field crews from AAFC; and university co-operative education students (Brittany Derrick, Russell Kirchner, Tom Zochowski, Chris Perra, Mirage Leung) and WS students (Graham Knibbs, Jeff Kerkovius, Stephen Kimanzi, Dan Lang, Leo Kerrigan, Danielle Homer, Jeffrey van Santen, Caroline Hedge) and Robert Roskoden (German Academic Exchange Service internship student). References Abalos, D., Sanchez-Martin, L., Garcia-Torres, L., Van Groenigen, J.W., Vallejo, A., 2014. Management of irrigation frequency and nitrogen fertilization to mitigate GHG and NO emissions from drip-fertigated crops. Sci. Total Environ. 490, 880– 888. Beare, M.H., Gregorich, E.G., St-Georges, P., 2009. Compaction effects on CO2 and N2O production during drying and rewetting of soil. Soil Biol. Biochem. 41, 611– 621. Benoit, M., Garnier, J., Gilles, B., 2015. Temperature dependence of nitrous oxide production of a luvisolicsoil in batch experiments. Process Biochem. 50, 79–85. Box, G.E.P., Cox, D.R., 1964. An analysis of transformations. J. Roy. Statist. Soc. Ser. B 26, 211–252. Carlisle, E., Smart, D., Williams, L.E., Summers, M., 2010. California Vineyard Greenhouse Gas Emissions: Assessment of the Available Literature and Determination of Research NeedsCalifornia sustainable wine growing alliance publication, San Francisco, CA. . (accessed 14.07.07.) http://www. sustainablewinegrowing.org/docs/GHGreport.pdf. Chantigny, M.H., Angers, D.A., Kaiser, K., Kalbitz, K., 2008. Extraction and characterization of dissolved organic matter. In: Carter, M.R., Gregorich, E.G. (Eds.), Soil Sampling and Methods of Analysis, 2nd ed., Canadian Society of Soil Science. CRC Press Boca, Raton, FL, pp. 617–635. Cochran, V.L., Sparrow, E.B., Schlentner, S.F., Knight, C.W., 1997. Long-term tillage and crop residue management in the subarctic: fluxes of methane and nitrous oxide. Can. J. Soil Sci. 77, 565–570. Environment Canada. (2014). Canadian climate normals or averages, for Summerland and Penticton, BC. http://climate.weather.gc.ca/climate_normals/ results_1981_2010_e.html (accessed 14.07.07.). Environment Canada. (2014). Canadian climate data on-line customized search. Canada’s National Climate Archive, for Summerland and Penticton, BC. http:// www.climate.weatheroffice.gc.ca/advanceSearch/searchHistoricData_e.html (accessed 14.07.07). FAO, Food and Agriculture Organization of the United Nations Statistics Division. (2013). http://faostat3.fao.org/download/Q/QC/E (accessed 15.10.18.). Fentabil, M.M., Nichol, C.F., Neilsen, G.H., Hannam, K.D., Neilsen, D., Forge, T., Jones, M.D., 2016. Effect of micro-irrigation type, N-source and mulching on nitrous oxide emissions in a semi-arid climate: an assessment across two years in a Merlot grape vineyard. Agric. Water Manage. 171, 49–62. Fentabil, M.M., 2016. Water conservation management practices in vineyards and apple orchards: strategies for mitigating greenhouse gas emissions. Ph.D. Thesis. Department of Earth and Environmental Sciences and Physical Geography, University of British Columbia Okanagan.

251

Hannam, K.D., Neilsen, G.H., Forge, T., Neilsen, D., Losso, I., Jones, M.D., Nichol, C., Fentabil, M.M., 2015. Irrigation practices: nitrogen amendments and mulches affect nutrient dynamics in a young Merlot (Vitis vinifera L.) vineyard. Can. J. Soil Sci. 96, 23–36. Herrmann, A., Witter, E., 2002. Sources of C and N contributing to the flush in mineralization upon freeze-thaw cycles in soils. Soil Biol. Biochem. 34, 1495– 1505. Homyak, P.M., Yanai, R.D., Burns, D.A., Briggs, R.D., Germain, R.H., 2008. Nitrogen immobilization by wood-chip application: protecting water quality in a northern hardwood forest. For. Ecol. Manage. 255, 2589–2601. Hutchinson, G.L., Livingston, G.P., et al., 1993. Use of chamber systems to measure trace gas fluxes. In: Rolston, D.E. (Ed.), Agro-ecosystem Effects on Trace Gases and Climate Change. ASA Spec. Publ. 55, ASA, CSSA, and SSSA, Madison, WI, pp. 63–78. Kenward, M.G., Roger, J.H., 1997. Small sample inference for fixed effects from restricted maximum likelihood. Biometrics 53, 983–997. Koponen, H.T., Martikainen, P.J., 2004. Soil water content and freezing temperature affect freeze-thaw related N2O production in organic soil. Nutr. Cycl. Agroecosys. 69, 213–219. Laidlaw, J.W., 1993. Denitrification and nitrous oxide emission in thawing soil. M.Sc. Thesis. Department of Soil Science, University of Alberta, Edmonton, AB. Lin, S., Iqbal, J., Hu, R., Feng, M., 2010. N2O emissions from different land uses in midsubtropical China. Agric. Ecosyst. Environ. 136, 40–48. Lin, S., Iqbal, J., Hu, R., Ruan, L., Wu, J., Zhao, J., Wang, P., 2012. Differences in nitrous oxide fluxes from red soil under different land uses in mid-subtropical China. Agric. Ecosyst. Environ. 146, 168–178. Littell, R.C., Milliken, G.A., Stroup, W.W., Wolfinger, R.D., Schabenberger, O., 2006. Analysis of repeated measures data, SAS for Mixed Models. second ed. SAS Institute, Cary, N.C, pp. 159–203. Livesley, S.J., Dougherty, B.J., Smith, A.J., Navaud, D., Wylie, L.J., Arndt, S.K., 2010. Soilatmosphere exchange of carbon dioxide, methane and nitrous oxide in urban garden systems: impact of irrigation fertiliser and mulch. Urban Ecosyst. 13, 273–293. Lopez-Fernández, S., Diez, J.A., Hernaiz, P., Arce, A., Garcia-Torres, L., Vallejo, A., 2007. Effects of fertiliser type and the presence or absence of plants on nitrous oxide emissions from irrigated soils. Nutr. Cycl. Agroecosyst. 78, 279–289. Millar, N., Robertson, G.P., Diamant, A., Gehl, R.J., Grace, P.R., Hoben, J.P., 2012. Methodology for Quantifying Nitrous Oxide (N2O) Emissions Reductions by Reducing Nitrogen Fertilizer Use on Agricultural CropsAmerican carbon registry, Winrock international, Little Rock, Arkansas. . (accessed 15.11.11.) http://americancarbonregistry.org/carbon-accounting/standardsmethodologies/emissions-reductions-through-reduced-use-of-nitrogenfertilizer-on-agricultural-crops/msu-epri-methodology-acr-v1-0_final.pdf. Neilsen, C.B., Groffman, P.M., Hamburg, S.P., Driscoll, C.T., Fahey, T.J., Hardy, J.P., 2001. Freezing effects on carbon and nitrogen cycling in northern hardwood forest soils. Soil Sci. Soc. Am. J. 65, 1723–1730. Neilsen, D., Van Der Gulik, T., Cannon, A., Taylor, B., Fretwell, R., 2015. Modeling future water demand for current and future climate in the Okanagan basin B.C, Canada. Acta Hort. 1063, 211–218. Nyamadzawo, G., Shi, Y., Chirinda, N., Olesen, J.E., Mapanda, F., Wuta, M., Wu, W., Meng, F., Oelofse, M., de Neergaard, A., Smith, J., 2014. Combining organic and inorganic nitrogen fertilisation reduces N2O emissions from cereal crops: a comparative analysis of China and Zimbabwe. Mitig. Adapt. Strateg. Glob. doi: http://dx.doi.org/10.1007/s11027-014-9560-9. Pang, J., Wang, X., Mu, Y., Ouyang, Z., Liu, W., 2009. Nitrous oxide emissions from an apple orchard soil in the semiarid Loess Plateau of China. Biol. Fertil. Soils 46, 37–44. Pelster, D.E., Chantigny, M.H., Rochette, P., Angers, D.A., Rieux, C., Vanasse, A., 2012. Nitrous oxide emissions respond differently to mineral and organic nitrogen sources in contrasting soil types. J. Environ. Qual. 41, 427–435. Rochette, P., Eriksen-Hamel, N.S., 2007. Chamber measurements of soil nitrous oxide flux: are absolute values reliable? Soil Sci. Soc. Am. J. 72, 331–342. Rochette, P., Hutchinson, G.L., 2005. Measuring soil respiration in situ: chamber techniques. In: Carter, M.R., Gregorich, E.G. (Eds.), Soil Sampling and Methods of Analysis, Canadian Society of Soil Science. CRC Press, Boca Raton, FL, pp. 851– 861. Rolston, D.E., Sharpley, A.N., Toy, D.W., Broadbent, F.E., 1982. Field measurement of denitrification: III. Rates during irrigation cycles. Soil Sci. Soc. Am. J. 46, 289– 296. Russow, R., Spott, O., Stange, C.F., 2008. Evaluation of nitrate and ammonium as sources of NO and N2O emissions from black earth soils (Haplic Chernozem) based on 15N field experiments. Soil Biol. Biochem. 40, 380–391. Sanchez-Martin, L., Meijide, A., Garcia-Torres, L., Vallejo, A., 2010. Combination of drip irrigation and organic fertilizer for mitigating emissions of nitrogen oxides in semiarid climate. Agric. Ecosyst. Environ. 137, 99–107. Seymour, R., 2015. Okanagan grape growers say apples the futureDaily Courier . . (accessed 15.10.15.) http://www.kelownadailycourier.ca/news/ article_c3fe38c4-51f8-11e5-8efa-fb1c06161fde.html. Shcherbak, I., Millar, N., Robertson, G.P., 2014. Global metaanalysis of the nonlinear response of soil nitrous oxide (N2O) emissions to fertilizer nitrogenPNAS 11 (25), 9199–9204. . (accessed 16.05.05.) http://www.pnas.org/content/111/25/ 9199. Smart, D.R., Alsina, M.M., Wolff, M.W., Matiasek, M.G., Schellenberg, D.L., Edstrom, J. P., Brown, P.H., Scow, K.M., 2011. N2O emissions and water management in California perennial crops. In: Guo, L., Gunasekara, A.M., McConnell, L.L. (Eds.),

252

M.M. Fentabil et al. / Agriculture, Ecosystems and Environment 235 (2016) 242–252

Understanding Greenhouse Gas Emissions from Agricultural Management. ACS Symposium Series, Am. Chem. Soci., Washington, DC, pp. 227–255. Smith, P., Daniel, M., Zucong, C., Daniel, G., Henry, J., Pushpam, K., Bruce, M., Stephen, O., Frank, O., Charles, R., Bob, S., Oleg, S., Mark, H., Tim, M., Genxing, P., Vladimir, R., Uwe, S., Sirintornthep, T., Martin, W., Jo, S., 2008. Greenhouse gas mitigation in agriculture. Phil. Trans. R. Soc. B Biol. Sci. 363 (1492), 789–813. Steenwerth, K.L., Belina, K.M., 2010. Vineyard weed management practices influence nitrate leaching and nitrous oxide emissions. Agric. Ecosyst. Environ. 138, 127–131. Van Groenigen, J.W., Kasper, G.J., Velthof, G.L., Dasselaar, A.V.D.P., Kuikman, P.J., 2004. Nitrous oxide emissions from silage maize fields under different mineral nitrogen fertilizer and slurry applications. Plant Soil 263, 101–111.

Venterea, R.T., Maharjan, B., Dolan, M.S., 2011. Fertilizer source and tillage effects on yield-scaled nitrous oxide emissions in a corn cropping system. J. Environ. Qual. 40, 1521–1531. Wang, F., Kang, Y., Liu, S., 2006. Effects of drip irrigation frequency on soil wetting pattern and potato growth in North China Plain. Agr. Water Manage. 79, 248– 264. Wittneben, U., 1986. Soils of the Okanagan and Similkameen Valleys Ministry of Environment Technical Report 18British Columbia Soil Survey, Report 52, Victoria, BC. . (accessed 15.05.16.) http://sis.agr.gc.ca/cansis/publications/ surveys/bc/bc52/bc52_report.pdf. Zebarth, B.J., Rochette, P., Burton, D.L., 2008. N2O emissions from spring barley production as influenced by fertilizer nitrogen rate. Can. J. Soil Sci. 88, 197–205.