Mineral nitrogen dynamics in irrigated rice–wheat system under different irrigation and establishment methods and residue levels in arid drylands of Central Asia

Europ. J. Agronomy 47 (2013) 65–76

Contents lists available at SciVerse ScienceDirect

European Journal of Agronomy journal homepage: www.elsevier.com/locate/eja

Mineral nitrogen dynamics in irrigated rice–wheat system under different irrigation and establishment methods and residue levels in arid drylands of Central Asia Krishna Prasad Devkota a,∗ , Ahmad Manschadi b , John P.A. Lamers c , Mina Devkota d , Paul L.G. Vlek c a

CIMMYT International, Kathmandu, Nepal University of Natural Resources and Life Sciences Vienna, Department of Crop Sciences, Konrad–Lorenz Str. 24, 3430 Tulln, Austria c Center for Development Research (ZEF), University of Bonn, Germany d Post-Doctoral Fellow, CIMMYT International, Kathmandu, Nepal b

a r t i c l e

i n f o

Article history: Received 28 October 2012 Received in revised form 24 January 2013 Accepted 29 January 2013 Keywords: Dry seeded rice Alternate wet and dry irrigation N balance Alternative establishment methods Residue retention

a b s t r a c t Increasing water shortage and low water productivity in the irrigated drylands of Central Asia are compelling farmers to develop and adopt resource conservation technologies. Nitrogen (N) is the key nutrient for crop production in rice–wheat cropping systems in this region. Nitrogen dynamics of dry seeded rice(aerobic, anaerobic) planted in rotation with wheat (well drained, aerobic) can differ greatly from those of conventional rice cultivation. Soil mineral N dynamics in flood irrigated rice has extensively been studied and understood, however, the impact of establishment method and residue levels on this dynamics remains unknown. Experiments on resource conservation technologies were conducted between 2008 and 2009 to assess the impact of two establishment methods (beds and flats) in combination with three (R0, R50 and R100) residue levels and two irrigation modes (alternate wet and dry (AWD) irrigation (all zero till), and a continuously flooded conventional tillage (dry tillage)) with water seeded rice (WSR) on the mineral N dynamics under dry seeded rice (DSR)-surface seeded wheat systems. N balance from the top 80 cm soil layers indicated that 32–70% (122–236 kg ha−1 ) mineral N was unaccounted (lost) during rice cropping. The amount of unaccounted mineral N was affected by the irrigation method. Residue retention increased (p < 0.001) the unaccounted mineral N content by 38%. With AWD irrigation, the N loss was not different among dry seeded rice in flat (DSRF), dry seeded rice in bed (DSRB), and conventional tillage WSR. Under different irrigation, establishment methods and residue levels, unaccounted mineral N was mainly affected by plant N uptake and soil mineral N content. Major amounts (43–58%) of unaccounted mineral N from DSR field occurred between seeding and panicle initiation (PI). During the entire rice and wheat growing seasons, NH4 N consistently remained at very high levels, while, NO3 N remained at very low levels in all treatments. In rice, the irrigation method affected NH4 N content. Effect of residue retention and establishment methods were not significant on NH4 N and NO3 N dynamics in both crops and years. Further evidence of the continuously fluctuating water filled pore spaces (WFPS) of 64% and the microbial aerobic activity of 93% at the top 10 cm soil surface during rice growing season indicates soil in the DSR treatments was under frequent aerobic–anaerobic transformation, a conditions very conducive for higher amounts of N loss. In DSR treatments, the losses appeared to be caused by a combination of denitrification, leaching and N immobilization. When intending to use a DSR management strategies need to be developed for appropriate N management, irrigation scheduling, and residue use to increase mineral N availability and uptake before this practices can be recommended. © 2013 Elsevier B.V. All rights reserved.

1. Introduction

∗ Corresponding author. Tel.: +977 1 4269564. E-mail addresses: [email protected] (K.P. Devkota), [email protected] (A. Manschadi), [email protected] (J.P.A. Lamers), [email protected] (M. Devkota), [email protected] (P.L.G. Vlek). 1161-0301/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.eja.2013.01.009

In Central Asia, rice (Oriza sativa L.) is grown in rotation with wheat (Triticum aestivum L.) on approximately 241,000 ha (FAOSTAT, 2010) in the irrigated lowland areas of the Amu Darya and Syr Darya river basins (Gupta et al., 2009). Water-seeded rice (WSR) is the primary rice establishment practice in Central Asia (Ntanos, 2001; Devkota, 2011). This practice includes land preparation which consists of 3–5 dry tillage events using a chisel-plow

K.P. Devkota et al. / Europ. J. Agronomy 47 (2013) 65–76

conventional flooded rice–wheat systems, significant N loss has been reported in Japan (Kyaw et al., 2005), Bangladesh (Timsina et al., 2001) and India (Sadanandan and Mahapatra, 1973; Aulakh et al., 2001), while no loss has been reported in Indo-Gangetic Plains (Pathak et al., 2006). Improved knowledge on N dynamics and N balance are essential to increase the N use efficiency through minimizing N losses, determine optimal levels of N-fertilizer application, and develop appropriate crop management practices for rice–wheat systems under alternative establishment methods with residue retention. Thus, the objective of this study was to examine mineral N dynamics and balance in a DSR-zero-tillage wheat system under water-saving AWD irrigation, residue retention, and alternative establishment methods in Central Asia. 2. Materials and methods 2.1. Site description The field experiments were conducted in the 2008 and 2009ricegrowing seasons at the Cotton Research Institute in the Khorezm region (41◦ 32 12 N, 60◦ 40 44 E) located in north-western Uzbekistan on the left bank of the Amu Darya River within the transition zone of the Karakum and Kyzalkum deserts. The climate of the area is arid, with a long-term average annual rainfall of approximately 100 mm. But the potential evapotranspiration (1200 mm y−1 ) always greatly exceeds precipitation. The mean annual temperature is 13.4 ◦ C with a minimum in January/February (−7 ◦ C) and a maximum in June/July (40 ◦ C) for the last 39 years (Fig. 1). The soil at the experimental site is an irrigated alluvial meadow soil (Russian Classification) or arenosol, gleyic, calcaric, sodic (FAO, Classification), sandy loam to loamy sand in texture with high soil salinity (2–16 dS m−1 ECe 1:1 in 0–15 cm soil depth), shallow (0.5–2 m) and saline groundwater table (2–4 dS m−1 ). Initial soil chemical parameters indicated a medium to high soil mineral N, low total soil N (0.04–0.05%), low soil organic matter (0.4–0.8%), and a moderate range of exchangeable potassium (Table 1). Hydraulic conductivity of this soil is fairly high (0.3 m d−1 in the top 20 cm and over 0.8 m d−1 below 30 cm). At field capacity, the top 10 cm soil has a volumetric soil moisture content of 30%, which corresponds to the soil water tension of 10 kPa. 2.2. Experimental design and treatments To quantify the effects of AWD irrigation, alternative crop establishment systems, and residue retention on soil mineral N dynamics Wheat season

Rice season

40

Wheat season

20

30

15

20

10

10

5

0

Rainfall (mm)

and disc, followed by several passes with a triplane and roller to create a uniform, compact and levelled soil surface (Pittelkowa et al., 2012). Following germination, farmers keep a water depth of 15–20 cm standing water throughout the rice-growing period, which amounts to more than 5000 mm of applied irrigation water (Devkota, 2011). Thus, excessive and inefficient water use (Gupta et al., 2009), intensive soil tillage, and diminishing supply of irrigation water in the region (Christmann et al., 2009) are compelling farmers to develop alternative rice establishment systems based on innovative technologies and management practices. In recent years, there has been a shift from transplanting to dry seeded rice (DSR) in Southeast Asia (Pandey and Velasco, 2002), where zero tillage DSR using alternate wet and dry (AWD) irrigation followed by zero tillage wheat are being adopted to reduce irrigation water input (Sudhir-Yadav et al., 2011a,b), reduce labour and input costs (Pandey and Velasco, 1999; Jat et al., 2009; Saharawat et al., 2010), improve water- and nutrient-use efficiency and avoid the deleterious effects of puddling and intensive soil tillage on soil structure and fertility (Timsina and Connor, 2001), and alleviate soil degradation problems (Ladha et al., 2009b; Farooq et al., 2011). While DSR with AWD irrigation have the potential to save irrigation water (Sudhir-Yadav et al., 2011b), they also alter soil aeration and thus the soil bio-geochemical processes, which in turn increase the loss of soil organic matter, reduce biological nitrogen fixation, and reduce plant-available N supply (Buresh and Haefele, 2010). Nitrogen dynamics have extensively been studied in continuous flood-irrigated rice (Buresh and De Datta, 1991; George et al., 1992), but limited information is available for DSR followed by zero-tillage wheat. In flood-irrigated rice with the soil being mostly under anaerobic conditions, N is available as ammonium nitrogen (NH4 N) (Vlek and Byrnes, 1986), whereas in DSR N is available mostly in the form of nitrate nitrogen (NO3 N) (George et al., 1992). AWD irrigation favours rapid fluctuation of soil moisture, soil redox potential, and water-filled pore spaces (WFPS) (Linn and Doran, 1984; Firestone and Davidson, 1989; Bouwman, 1996; Scholefield et al., 1997), which enhance soil aeration and ultimately lead to the formation of NO3 N (Patrick and Wyatt, 1964; Reddy and Patrick, 1976; George et al., 1992; Johnson-Beebout et al., 2009). Since conversion of NO3 to NH4 is negligible, the accumulated NO3 N is prone to losses either by denitrification or leaching upon soil flooding (Buresh et al., 1989; George et al., 1992, 1993). Reduced tillage and increased maintenance of crop residues on the soil surface often results cooler, wetter, and more compacted than conventionally tilled soil (Mielke et al., 1986). These differences in soil physical environment largely influence WFPS, which affects the associated N transformation processes (Linn and Doran, 1984). No-tillage with crop residue retention induces higher microbial (urease) activity on the soil surface through nitrification and denitrification, which increase N2 O emissions (Granli and Bockman, 1994) or volatilization (Al-Fiananir and Maclftnzie, 1992; Rochette et al., 2009; Dusserrea et al., 2012). But appropriate water and residue management can reduce N loss by emission (Johnson-Beebout et al., 2009). Intensive soil tillage accelerates N mineralization of crop residues and soil organic N (Sainju and Singh, 2001) and increases accumulation of NO3 N in the soil profile (Al-Kaisi and Licht, 2004). Besides N2 O emission, the increased accumulation of NO3 N in the soil profile increases the potential for N leaching to shallow water tables (Keeney and Follett, 1991) or crop residue may immobilize N (Quemada and Cabrera, 1995). Thus, availability and dynamics of soil mineral N (NO3 and NH4 N) may differ also with tillage method and residue retention. Also, partial N balance of dry seeded rice-(aerobic, anaerobic) planted in rotation with wheat (well drained, aerobic) may differ substantially from that of conventional flooded rice–wheat systems because of differences in soil and water management practices. In

Temperature (0C) and -2 -1 solar radiation (M J m d )

66

0 Max. Temperature Min. Temperature

-10 0

50

100

150

200

Solar radiation Rainfall

250

300

-5 350

Day of a year ◦

Fig. 1. Daily air temperature ( C) and solar radiation (M J m−2 d−1 ) and monthly total rainfall (mm) at experimental site, Urgench, Khorezm, Uzbekistan during rice and wheat growing period (from data from 1970 to 2009).

K.P. Devkota et al. / Europ. J. Agronomy 47 (2013) 65–76

67

Table 1 Initial soil properties of the experimental site in 2008. Depth (cm)

Bulk density (g cm −3 )

Soil pH

NH4 N (mg kg−1 )

NO3 N (mg kg−1 )

Total N (%)

Soil organic carbon (%)

Available phosphorus (mg kg −1 )

Exchangeable potassium (mg kg −1 )

0–10 10–20 20–30 30–60 60–90

1.35 1.41 1.42 1.52 1.57

5.57 5.56 5.57 5.69 5.78

5.4 6.5 6.3 6.3 5.2

5.3 4.4 5.2 4.0 3.9

0.05 0.05 0.04 0.03 0.03

0.36 0.30 0.26 0.23 0.19

27.9 25.9 21.9 19.2 17.6

98.5 95.0 89.3 81.4 76.8

Note: Soil organic matter = 1.56 × Soil organic carbon.

in rice–wheat system, seven treatments in 2008 and eight in 2009 were compared. Since the continuously flooded control plots would raise the groundwater level and interfere with the AWD treatments, these plots could not be randomly allocated within the experimental blocks. Therefore, a non-experimental design was used (Wludyka, 2011) in which the AWD treatments were laid out in a randomized block design 70 m from the flood-irrigated control treatment. The individual plot size was 480 m2 . Treatments included the combinations of two establishment methods (beds and flats) with three residue levels with AWD (all zero till DSR), and a continuously flooded conventional tillage (dry tillage) with water seeded rice (WSR). In the second year, a WSR treatment with AWD (WSRF-R0-AWD) was added (Table 2). The three levels of residue retention included (i) no residue retention (R0); crop harvested from the base leaving 3–5 cm stubble according to farmers’ practice, (ii) R50; retaining 15–20 cm straw stubble, and (iii) R100; retaining 35–40 cm straw stubble representing the retention of a maximum possible amount of residue. At the onset of the experiment in 2008, crop residues from the previous crop were not available. Therefore, chopped wheat residues were imported and uniformly spread over the plot soil surface in the R50 and R100 treatments at 1.5 and 3.0 t ha−1 , respectively. In 2009, standing residues from rice 2008 and wheat 2009 were retained; at the time of sowing the 2nd rice crop, the cumulative crop residues amounted to 8.5 t ha−1 in the R50 and 14.3 t ha−1 in the R100 treatment (Table 3). The six AWD rice treatments, i.e. two establishment methods combined with three levels of residue retention, and the added treatment in 2009 (WSRF-R0-AWD), were irrigated almost daily during the first 15 days after sowing (DAS). Thereafter, AWD rice was flood-irrigated when the volumetric soil moisture content at 20 cm soil depth dropped 5–10% below the field capacity, i.e. at a soil water tension of around 20 kPa, i.e. the field plots were irrigated at 1–5 day intervals until one week before rice harvest. WSRF-R0FI rice was irrigated twice a day to maintain a 5–15 cm standing

water level, which corresponds to farmers’ practice of irrigation in the region until one week before crop harvest. In WSR treatments, 24-h water-soaked pre-germinated rice seed was uniformly broadcasted into the standing water. In DSRF treatments, rice was seeded in 20 cm spacing using a tractor drawn seed-cum-fertilizer zero-till drill. In DSRB treatments, rice was planted on beds (37 cm wide at the top, 15 cm height, and 30 cm wide furrows). Two rows of rice were direct-seeded on each raised bed at 20 cm row spacing. As all plots were ploughed and fertilizer was incorporated into the soil before seeding, there was not a real zero tillage treatment in 2008. 2.3. Cultivation practices A short duration (90–100 days) high yielding and inputresponsive local rice variety (Nukus-2) was seeded on 18 June in 2008 and on 21 June in 2009 (after wheat harvest) using a seed rate of 140 kg ha−1 in the DSR treatments. At the same time and rate, pre-soaked rice seeds were seeded in WSRF-R0-FI and WSRF-R0-AWD. Fertilizer was applied at 257 kg N ha−1 in 2008 and 250 kg N ha−1 in 2009, and 120 kg P2 O5 and 80 kg K2 O ha−1 in both years. In 2008, all P2 O5 and K2 O fertilizers were incorporated prior to planting and 80 kg N ha−1 in AWD and 29 kg N ha−1 in WSRF-R0FI were drilled at the time of seeding. The remaining N was applied in four splits shortly after irrigation. In 2009, all P2 O5 and K2 O were broadcast 30 days after seeding, while N was applied in four splits (Table 3). In 2008, following the establishment of DSRB and DSRF plots, two pre-sowing irrigations were applied to allow weeds to germinate. Next, all weeds were treated with Glyphosate at the rate of 2 ml l−1 water 2–3 days before rice seeding. To control postemergence weeds, Gulliver (Azimsulfuron 50 WG) at 25 g ha−1 was applied 28 DAS in both years. In addition, weeds were also removed manually 4 and 7 weeks after emergence. The wheat variety Krasnodar-99 was surface-broadcast seeded using a seed rate of 200 kg ha−1 into the standing rice in all

Table 2 Treatment details of the rice–wheat experiment in the Khorezm region of Uzbekistan 2008–2009. Treatment

Rice

Wheat

DSRB-R0-AWD (T1)

Dry seeded rice grown on bed, no residue retention, alternate wet and dry irrigation, fresh beds, conventional tillage (CT) in 1st year, zero-tillage (ZT) in 2nd year, drill seeded in both years Dry seeded rice on bed, 50% residue retention, alternate wet and dry irrigation, fresh beds, CT in 1st year, ZT in 2nd year, drill seeded in both years Dry seeded rice on bed, 100% residue retention, alternate wet and dry irrigation, fresh beds, CT in 1st year, ZT in 2nd year, drill seeded in both years Dry seeded rice on flat, no residue retention, alternate wet and dry irrigation, CT in 1st year, ZT in 2nd year, drill seeded in both years Dry seeded rice on flat, 50% residue retention, alternate wet and dry irrigation, CT in 1st year, ZT in 2nd year, drill seeded in both years Dry seeded rice on flat, 100% residue retention, alternate wet and dry irrigation, CT in 1st year, ZT in 2nd year, drill seeded in both years Water seeded rice on flat, no residue retention, conventional flood irrigation (FI), CT in both years Water seeded rice on flat, no residue retention, alternate wet and dry irrigation, CT, new treatment in 2nd year

Surface seeded on permanent beds in standing rice, residue harvested

DSRB-R50-AWD (T2) DSRB-R100-AWD (T3) DSRF-R0-AWD (T4) DSRF-R50-AWD (T5) DSRF-R100-AWD (T6) WSRF-R0-FI (T7) WSRF-R0-AWD (T8)

Surface seeded in permanent beds in standing rice, 50% residue retention Surface seeded in permanent beds in standing rice, 100% residue retention Surface seeded in zero-till flat in standing rice, residue harvested Surface seeded in zero-till flat in standing rice, 50% residue retention Surface seeded in zero-till flat in standing rice, 100% residue retention Surface seeded into standing rice Surface seeding into standing rice

4146 6689

– 17 DAS 30 kg N ha−1 30 DAS 79:120:80 kg N:P2 O5 :K2 O ha−1 45 DAS 100 kg N ha−1 59 DAS 41 kg N ha−1

21 June 22 October 140 250:120:80

K.P. Devkota et al. / Europ. J. Agronomy 47 (2013) 65–76

Rice 2009

68

treatments at 17 days before rice harvest (Table 3). In wheat, fertilizer dose of 124:100:70 kg NPK were applied and phosphorus and potassium fertilizers were broadcast applied in the first week of November, while N was applied in two equal splits in March and April; N was applied as urea granules (Table 3). Weed control was carried out through the application of the herbicide Granstar 75 DF (75% Tribenuron methyl) at the rate 25 g ha−1 in the last week of March. Wheat was irrigated eight times during the crop-growing period, i.e., two times before fertilizer application and at the major growth stages, i.e., at emergence, spike initiation, booting, flowering, milking and soft dough.

– 39:100:70 kg N:P2 O5 :K2 O ha−1 on 5 November 2009 42.5 kg N ha−1 on 15th March 2009 42.5 kg N ha−1 on 15th April 2009 108:120:80 kg N:P2 O5 : K2 O ha−1 in AWD and 28:120:80 kg N:P2 O5 :K2 O ha−1 in WSR 16 DAS 20 kg N ha−1 34 DAS 43 kg N ha−1 in AWD and 123 kg N ha−1 in WSR 53 DAS 43 kg N ha−1 64 DAD 43 kg N ha−1

2922 4657

23 September 2008 13 June 2009 200 124:100:70 18 June 8 October 140 257:120:80

Seeding Harvesting Seed rate (kg ha−1 ) Fertilizer dose (NPK kg ha−1 ) Fertilizer application Basal 1st topdressing 2nd topdressing 3rd topdressing 4th topdressing Residue level (kg ha−1 ) Residue in R50 Residue in R100

Initial soil samples were collected before the start of the experiment to determine the fertility status of the experimental site (Table 1). Both plant and soil samples were collected at major growth stages, i.e., at emergence (15 days after seeding; DAS), panicle initiation (45 DAS), flowering (75 DAS) and physiological maturity stages in rice and at emergence (15 DAS), spike initiation (182 DAS), flowering (231 DAS) and physiological maturity stages in wheat. Soil sample at rice harvest is the initial for wheat and vice versa. Soil samples for the amount of mineral remained in the soil was collected on the same day of plant sampling for N uptake. Plants samples for the determination total aboveground N uptake were collected from a 0.8 m2 area. Root, stems, green leaves, senesced leaves and panicles were separated and oven dried separately for 72 h at 65 ◦ C until constant weight. Total plant N was analyzed by the Kjeldahl method (Bremner and Mulvaney, 1982). In both crops soil samples were collected from three different points of each plot from 0–10, 10–20, 20–30, 30–50 and 50–80 cm soil depths; the samples were composited. Soil NO3 N content (mg kg−1 ) was examined using the Granvald–Ljashu method, while the NH4 N content (mg kg−1 ) was determined by colorimetric analysis using the Nessler reagent (Protasov, 1977). The NH4 and NO3 N contents (kg ha−1 ) at a particular depth were determined by multiplying N concentrations with the respective bulk density. 2.5. Groundwater nitrate The nitrate content in the groundwater was measured before and after irrigation and fertilizer application during the entire cropgrowing period in 2009. A total of 15 piezometers (3 in the WSR and 12 in the DSR treatments) were installed randomly in down to 2.5 m soil depth. Water samples were collected from each piezometer and analyzed immediately for NO3 N concentration test using procedure as described by Merck (2013). 2.6. Partial N balance

1500 3000

Table 3 Time of seeding and harvesting and input use in rice–wheat system 2008–2009.

Wheat 2009 Rice 2008 Item

2.4. Plant and soil sampling and analysis

Mineral N balance was determined in two steps, i.e., at harvest of each crop and at major crop growth stages (panicle initiation, flowering, and physiological maturity for rice and spike initiation, flowering, and physiological maturity for wheat). Partial N balance at crop harvest was determined based on the total soil mineral N content before crop seeding, soil mineral N remained at crop harvest, total N uptake by the crop, and total amount of N added from fertilizer during the crop growing period. Similarly, partial N balance at different crop growth stages was determined based on the total soil mineral N content in the earlier sampling date (initial soil mineral N), total mineral N remained at the sampling date, N added from fertilizer, and N uptake by the crop from earlier to this sampling date. Total plant N uptake was determined from the sum of root, leaf, stem, senesced leaf, and panicle by multiplying dry biomass with N concentration (%) of the respective plant part. The N input was calculated by adding the initial soil mineral N content

K.P. Devkota et al. / Europ. J. Agronomy 47 (2013) 65–76

at 0–80 cm soil depth to the amount of N from added N fertilizer (Liu et al., 2003). Similarly, N output was calculated by adding plant N uptake to the mineral N content remained in the soil. The difference between N input and output is considered as unaccounted mineral N or N possibly lost. Nitrogen not measured directly includes N input from rainfall and irrigation water, N mineralization from the organic N pool, gaseous N losses from ammonia volatilization and denitrification and leaching, N immobilized by the crop residue, and N released from residue decomposition.

ThetaProbe ML2X FDR sensors. Sensor data were validated using the volumetric moisture content measured manually at 10 sampling dates in rice 2008, 23 sampling dates in rice 2009, and 11 sampling dates in wheat 2009 in the respective soil depths. The water-filled pore space was calculated using the measured volumetric moisture content, soil porosity, bulk density, and particle density of the soil using the equation given by Linn and Doran (1984); WFPS (mL mL−1 ) =

2.7. Water-filled pore space and relative aerobic activity Water-filled pore space (WFPS) was calculated from the volumetric moisture content of soil measured at 10, 30, 50 and 70 cm depths. Volumetric moisture content was measured at hourly intervals during the entire crop growing period in two replications by

69

Porosity (%) =



1−

Percent volumetric water content × 100 Percent total soil porosity Bulk density Particle density



× 100

A soil particle density of 2.65 g cm−3 was used when calculating porosity (Scheer et al., 2008). Similarly, relative aerobic activity in

Fig. 2. Nitrogen from fertilizer, initial soil mineral N, plant N uptake, and mineral N remained in soil at harvest (kg ha−1 ) in rice 2008 (A), rice 2009 (B) and wheat 2009 (C) as affected by different irrigation and establishment methods and residue levels. LSD (0.05) is least significant difference between treatments. For detail treatment description see Table 2.

70

K.P. Devkota et al. / Europ. J. Agronomy 47 (2013) 65–76

different soil depths during the rice and wheat growing period was calculated using the equation given by Linn and Doran (1984) for WFPS higher than 60% as:

3.1. Partial N balance in rice–wheat system

WSRF-R0-FI followed by WSRF-R0-AWD and DSRF-R0-AWD had the lowest amount of unaccounted mineral N. The unaccounted mineral N was mainly affected (p < 0.05) by the plant N uptake in 2008, and by both plant N uptake and soil mineral N content in 2009. Both plant N uptake and soil mineral N content were significantly lower in DSRF-R100-AWD than in the other treatments in 2009. At wheat harvest, unaccounted mineral N was not observed, and the averaged N output (243 kg ha−1 ) was higher by 6% than input (229 kg ha−1 ) (Fig. 2C). A residual effect of AWD treatments of rice was observed in wheat, where N output was higher by 8% in the DSR than under WSR. Among the DSR treatments, there was no significant difference in N input and output. In rice–wheat system, combined over two rice seasons and one season of wheat, the total amount of unaccounted mineral N was significantly higher under DSRF-R100-AWD (366 kg ha−1 ) followed by DSRB-R100-AWD (346 kg ha−1 ) and the lowest under WSR (250 kg ha−1 ). In the rice–wheat system, mineral N dynamics were affected by irrigation method, and the unaccounted mineral N content was higher under R0 treatments of DSR than under WSR by 10% (p = 0.05). It was also affected by residue retention, and was higher by 58% (p < 0.001) under R100 than under R0 treatments of DSR, while the N losses were same under R0 AWD treatments of DSRB, DSRF and WSR establishment methods.

3.1.1. At crop harvest At rice harvest, 32–43% (122–163 kg ha−1 ) in 2008 (Fig. 2A) and 38–70% (125–236 kg ha−1 ) in 2009 (Fig. 2B) of mineral N was unaccounted (lost). The unaccounted mineral N was affected by the irrigation method, and was higher (p < 0.05) in the R0 AWD treatments, by 11% in 2008 and 29% in 2009, than under WSRF-R0-FI. This is further supported by 27 kg ha−1 (21%) higher unaccounted mineral N in WSRF-R0-AWD compared to WSRF-R0-FI in 2009. The unaccounted mineral N was also affected by residue retention as it increased (p = 0.05) by 10% in 2008 and 28% in 2009 under R100 compared to R0 treatments of DSR. No significant differences in unaccounted mineral N among R0 AWD treatments of DSRB, DSRF and WSRF indicate that the establishment method did not affect N losses. Among the different establishment methods in 2009, DSRFR100-AWD followed by DSRB-R100-AWD had the highest, while

3.1.2. At different crop growth stages In rice 2008, the largest amount of mineral N (43–58%; 127–169 kg ha−1 ; Fig. 3A) was unaccounted in the rice field from seeding to panicle initiation (PI) followed by (13–33%; 22–64 kg ha−1 , Fig. 3B) from PI to flowering stages. From seeding to PI (45 days after seeding) stage, unaccounted mineral N was affected (p < 0.05) by irrigation method and treatments of R0 AWD had 23% higher unaccounted mineral N than WSR (Fig. 3A), however, it was same among DSR treatments. At both measurements, N output was significantly higher in the WSR than in the DSR treatments, mostly due to higher (p = 0.06) N uptake. Total mineral N content in the 0–80 cm soil profile was not significantly different among treatments at both stages (Fig. 3A). In rice 2009, unaccounted mineral N content was affected by irrigation method and residue level (Fig. 4). In R0 AWD treatment,

Relative aerobic activity =

60% Percent water filled pore space

2.8. Statistical analysis The non-experimental design was analyzed as an analyses of variance using PROC GLM (SAS-Institute, 2008). Given the significant year × treatment interaction effect (p < 0.05), data were analyzed and presented year-wise. To quantify the main effect of establishment methods, residue levels, and irrigation methods, orthogonal single degree of freedom contrasts were used. Repeated measure analysis was performed for those observations recorded over time. Differences between individual treatments were analyzed using Fisher’s Protected Least Significant Difference (LSD). Treatments differences were considered statistically significant at p < 0.05. 3. Results

Fig. 3. Nitrogen from fertilizer, initial soil mineral N, plant N uptake, and mineral N remained in soil (kg ha−1 ) from seeding to PI (A) and PI to flowering (B) stages of rice as affected by different irrigation and establishment methods and residue levels in 2008. LSD (0.05) is least significant difference between treatments. For detail treatment description see Table 2.

K.P. Devkota et al. / Europ. J. Agronomy 47 (2013) 65–76

71

Fig. 4. Nitrogen from fertilizer, initial soil mineral N, plant N uptake, and mineral N remained in soil (kg ha−1 ) from seeding to PI (A) and PI to flowering (B) stages of rice as affected by different irrigation and establishment methods and residue levels in 2009. LSD (0.05) is least significant difference between treatments. For detail treatment description see Table 2.

more than half of the total input mineral N was unaccounted from seeding to PI, in contrast, at the same stage input and output were not different (p > 0.05) in WSR. At PI to flowering stage, R0 AWD treatment had 55% higher (p < 0.001) unaccounted mineral N than in WSR. Further, it was also supported by the higher (p < 0.05, 24%) unaccounted mineral N in WSRF-R0-AWD than in WSR. It was also affected by residue retention and R100 had 35 and 25% higher unaccounted mineral N than in R0 treatments of DSR from seeding to PI and PI to flowering stages, respectively. In the AWD treatments, unaccounted mineral N did not differ among DSRB, DSRF and WSR at all growth stages. In wheat, at most of the growth stages N output was at par with N input except at spike initiation to booting (Fig. 5), where, on an average 29 kg mineral N ha−1 found unaccounted. A residual effect of the DSR treatments of rice was observed in wheat on soil

mineral N content and N uptake. Total soil mineral N content was higher by 7% from seeding to spike initiation and by 19% from spike initiation to flowering, in DSR treatments than under WSR (Fig. 5). R100 treatments of DSR had higher mineral N availability by 10% from seeding to spike initiation and by 21% from spike initiation to flowering than R0. 3.1.3. N uptake in rice and wheat crops In rice, the total N accumulation in aboveground biomass at physiological maturity ranged from 104 to 145 kg ha−1 in 2008 (Fig. 2A) and 52–139 kg ha−1 in 2009 (Fig. 2B). At most of the sampling dates N accumulation was consistently higher (p < 0.05) in the WSR than in the DSR treatments (Figs. 3 and 4). Irrespective of establishment method, R0 treatments had consistently higher N uptake followed by R50 and R100 treatments of DSR in both years;

Fig. 5. Nitrogen from fertilizer, initial soil mineral N, plant N uptake, and mineral N remained in soil (kg ha−1 ) from seeding to spike initiation (A), spike initiation to booting (B) stages of wheat as affected by different establishment method and residue level 2009. LSD (0.05) is least significant difference between treatments. For detail treatment description see Table 2.

K.P. Devkota et al. / Europ. J. Agronomy 47 (2013) 65–76

NH4- and NO3-N (mg kg-1)

72

10-20

0-10 20-30 50-80

A 10 8 6

C

B

30-50

10

LSD (0.05)

10

LSD (0.05) NH4-N NO3-N

8

NH4-N NO3-N

8

6

4

4

2

2

6 4 SI

0

FL

PI

E

0

20

40

60

PM

80

LSD (0.05) NH -N 4 NO3-N

FL PM

FL

PI

E

2

PM

E 0

0

100

50

Days after sowing

100

150

200

250

0

20

40

Days after sowing

60

80

100 120

Days after sowing

Fig. 6. Soil NH4 and NO3 N concentrations (mg kg−1 ) at different soil depths during rice 2008 (A), wheat 2009 (B) and rice 2009 (C) growing seasons. E = emergence, PI = panicle initiation/SI = spike initiation, FL = flowering, PM = physiological maturity stages. Dotted lines are the NO3 N and the solid lines are NH4 N. Lines are connected for better visualization.

it was significantly higher in 2009. In wheat, total aboveground plant N accumulation at physiological maturity ranged from 132 to 158 kg ha−1 but the differences were statistically not significant (Fig. 2C). However, there were some significant differences at spike initiation and booting stages (Fig. 5). The non-significant difference in N uptake at physiological maturity in R0 AWD treatments indicated no effect of no-till bed and flat and conventional tillage establishment methods in both crops and years (Fig. 2).

than under R0 treatments of DSR by 18% in 2008 and by 44% in 2009 at tillering (Fig. 7B). After tillering, NH4 N was not significantly different among residue levels in both years. The NH4 N concentration was higher in the WSR than in the DSR treatments in all soil depths at major crop growth stages (tillering to flowering) (data not shown). In wheat, NH4 N content in the top 80 cm soil profile was not affected by establishment method and residue treatments at all growth stages (Fig. 7). Averaged NH4 N content was higher (p = 0.06) under DSR than under WSR. NH4 N content was higher (p < 0.05) under R100 than under R0 treatments of DSR by 11% at booting, and by 27% at flowering.

3.1.4. Soil mineral nitrogen dynamics in the rice–wheat system In both the rice and wheat season, NH4 N and NO3 N concentration consistently decreased (p < 0.001) from 0- to 80-cm soil depth in all treatments (Fig. 6).

3.1.6. NO3 N dynamics In rice 2008, soil NO3 N content was affected (p < 0.05) by irrigation method at early crop growth stage (Fig. 7A). The R0 treatments of DSR showed 46% higher NO3 N than WSR at PI. At this stage, DSRB-R0-AWD had 43% higher NO3 N content than DSRFR0-AWD. However, in rice 2009, NO3 N content was not affected by irrigation, establishment method and residue level at all growth

NH4- and NO3-N (kg ha-1)

3.1.5. NH4 N dynamics In rice, NH4 N content in the top 80 cm soil profile was affected (p < 0.05) by irrigation method, and was higher under WSR by 26 and 32% at PI and 8 and 24% at flowering in 2008 and 2009, respectively, than in R0 treatments of DSR (Fig. 7A). NH4 N was also affected by residue level, and the values were higher under R100 A

120

Rice 2008

120 DSRB-R0-AWD DSRF-R0-AWD WSRF-R0-FI

100 80 60

A

100

100

80

80

0 0

20

40 NO3-N

40

60

NH4-N

80

100 120

LSD (0.05)

40

NO3-N

0 180

NH4-N

0

200

220

240

0

260

20

40 60 80 100 120 Days after sowing

NH4 and NO3-N (kg ha-1)

Days after sowing

B

120

Rice 2008

100 80 60

DSRB-R0-AWD DSRB-R50-AWD DSRB-R100-AWD DSRF-R0-AWD DSRF-R50-AWD DSRF-R100-AWD

120

B

LSD (0.05) NH4-N

20

100

100

80

80

0 0

20

40

40 NO3-N 60

20 80

Days after sowing

B

120

Wheat 2009

60

40

NO3-N

20

20

Days after sowing

WSRF-R0-AWD

LSD (0.05)

40 LSD (0.05) NH4-N

Rice 2009

A

60

60

20

120

Wheat 2009

Rice 2009

60 LSD (0.05) NH4-N

LSD (0.05) NH4-N NO3-N

40

NO3-N

20

100 120

0 180

0 200

220

240

Days after sowing

260

20

40

60

80

100

120

Days after sowing

Fig. 7. Total NO3 and NH4 N content (kg ha−1 ) in top 80 cm soil profile under different irrigation and establishment methods (A) and residue levels (B). Solid lines indicate NH4 N and dotted lines NO3 N. LSD (0.05) is least significant difference between treatments over time. Lines are connected for better visualization.

K.P. Devkota et al. / Europ. J. Agronomy 47 (2013) 65–76

NO3-N in groundwater (ppm)

DSR

WSR

16 14 12 10 8 6 4 2 0 0

20

40

60

80

100

120

Days after sowing Fig. 8. Groundwater NO3 N concentration (mg L−1 ) in WSR and DSR treatments, 2009. WSR = conventional tillage with flood irrigation, DSR = dry seeded rice with alternate wet and dry irrigation.

73

3.1.8. Water-filled pore space and relative aerobic activity During the rice growing season, soil pores under WSR were filled with water, thus the soil was continuously under anaerobic conditions. In the DSR treatments, averaged water-filled pore space (WFPS) was 64, 74, 88, and 104% in 10, 30, 50 and 70 cm soil depths, respectively; the soil in the top 30 cm was mostly under aerobic conditions (Fig. 9A and B). However, the continuously fluctuating WFPS during the rice season indicates that the soil in the DSR treatments was under frequent aerobic–anaerobic transformation. During the wheat season, WFPS was 57, 68, 79, and 95% in 10, 30, 50 and 70 cm soil depths, respectively (Fig. 9C). The calculated microbial aerobic activity under DSR treatments during the rice growing season was 93% in the top 10 cm soil followed by 82, 69 and 57% in 30, 50 and 70 cm soil depths, respectively. Similarly, during the wheat season, the microbial aerobic activity was 107, 89, 78 and 63% in 10, 30, 50 and 70 cm soil depths, respectively. 4. Discussion 4.1. Effect of irrigation method

stages (Fig. 7). The averaged NO3 N content at different growth stages were 12 kg ha−1 in the top 80 cm soil profile. Similarly, during the wheat season, NO3 N content in the soil profile was not significantly different among establishment methods and residue levels at all growth stages, and the averaged value was 12 kg ha−1 (Fig. 7).

10 30

A

B

50 70

Water-filled pore spaces (%)

Water-filled pore spaces (%)

3.1.7. Groundwater nitrate content Groundwater NO3 N concentration in DSR (1.86 + 0.05 mg l−1 ) treatments was higher by 52% than under WSR (1.22 + 0.07 mg l−1 ; Fig. 8).

Nitrogen depletes from rice field under frequent wet and dry irrigation. The higher unaccounted mineral N in DSR compared to WSR could be related to the low plant N uptake (Figs. 2–5); and low soil mineral N availability (Fig. 7) and higher N losses through various pathways. The lower N uptake under DSR with AWD irrigation could be due to slow early crop growth and development rates due to water stress in AWD treatments, which lead to reduced above- and belowground biomass accumulation and N demand (Devkota, 2011). This finding is in consistent with the finding of Belder et al. (2005) and Beyrouty et al. (1994). Further, due to two reasons the soil and water management conditions in AWD/DSR

100

80

60

100

80

60

40

40 0

20

40

60

80

100

0

120

20

Water-filled pore spaces (%)

40

60

80

100

120

Days after sowing

Days after sowing

C 100

80

60

40 0

20

40

60

200

250

Days after sowing Fig. 9. Water-filled pore spaces (%) at different soil depths (cm) in DSR treatments under rice–wheat system, 2008–2009. Rice 2008 (A), Rice 2009 (B) and Wheat 2009 (C).

74

K.P. Devkota et al. / Europ. J. Agronomy 47 (2013) 65–76

treatments shows frequent aerobic-anaerobic phases, a soil condition very conducive for significant amount of N loss through nitrification–denitrification and/or ammonia volatilization and/or leaching: (1) During the crop growing period, 1700–2200 mm irrigation water in 40 irrigation events in DSR and 5900−6700 mm in more than 70 irrigation events in WSR (data not shown) were applied (Devkota, 2011). The frequent application of 4–5 cm flood irrigation indicates the occurrence of temporary anaerobic conditions for 5–6 h (as this amount of water generally disappeared within 4–5 h after application), while, the re-application in 20 kPa at 1–5 days interval (Section 2.2) indicates the temporary aerobic soil conditions. However, as the field was continuously flooded, soil was mostly under anaerobic condition in continuously flooded WSR. (2) The continuously and frequently fluctuating WFPS of 57–82% in the top 30 cm soil surface (Fig. 9) and 80% microbial aerobic activity (Section 3.5, Linn and Doran, 1984; Bateman and Baggs, 2005) further supports for frequent aerobic–anaerobic phases in DSR treatments. Frequent aerobic–anaerobic phases could change the soil redox potential frequently and thus the NO3 and NH4 N transformations, which influence ammonification, nitrification, and denitrification processes in the soil (Reddy and Patrick, 1976; Vlek and Byrnes, 1986). Significant N losses occur through N2 O emission in WFPS of 60–70% (Linn and Doran, 1984; Bateman and Baggs, 2005). In a study conducted in the Khorezm region, (Scheer et al., 2008) reported high N2 O losses (more than 50 ␮g N2 O N m−2 h−1 ) at a WFPS greater than 60%, while lower losses (<10 ␮g N2 O N m−2 h−1 ) were measured under continuously floodirrigated condition. Higher amount of N loss in DSR than in WSR at early growth stages of rice (Figs. 3–4) were due to faster growth rate and higher N uptake in WSR than in DSR (Devkota, 2011). The higher NO3 N concentration in the groundwater in the DSR than in the WSR treatments in 2009 (Fig. 8) indicates that a considerable amount of N could also have been lost through leaching. Below a depth of 45 cm, the soil at the study site mostly consists of sand and has a high infiltration rate of 0.5–1.0 cm h−1 (Devkota, 2011). Rapid loss of irrigation water (4–5 cm applied water within 5–6 h) is a situation conductive to NO3 N leaching (Bergstrom and Johansson, 1991). Alternate wetting and drying is associated with increased NO3 N during the aerobic phase (Linn and Doran, 1984; George et al., 1993). As frequent wet and dry irrigation was applied in DSR treatments, the accumulated NO3 N during the aerobic phase could have been lost through leaching during the subsequent irrigation. Thus, although the actual amount of N loss through leaching was not monitored, due to the high permeability of the soil layer and the application of frequent dry and wet flood irrigation, a significant amount of unaccounted mineral N in the DSR treatments could have been lost through the leaching of NO3 N. NH4 N in soil profile remained significantly higher in the WSR than in the DSR treatments during the active crop growth stages, i.e. from panicle initiation to flowering in both years (Fig. 7). The applied N remained adsorbed in the form of NH4 N under floodirrigated conditions. NH4 N is the predominant and preferred form of N in flooded soils (George et al., 1992); it adsorbs on soil colloids, stays longer and leads to lower percolation loss (Cassman et al., 1998). There is also less possibility of NH4 N changing to NO3 N under flooded conditions (Vlek and Byrnes, 1986). In contrast, the consistently lower NH4 N in the DSR than in the WSR, and also the absence of differences in NO3 N concentration between the DSR and WSR irrigation methods, indicate that the applied N in the DSR treatments neither remained longer in the form of NH4 N nor in NO3 N in soil profile and possibly could have been lost. The NO3 N during rice growing season 2009 was consistently low due to (1) low NO3 N content in soil at the time of rice seeding and N fertilizer was applied only after 27 DAS, (2) there could have been a greater flux of NO3 N in the soil but it may not have been captured in the soil samples, as these had been collected only at

3-week intervals and always before fertilizer application, and (3) due to the frequent wetting and drying irrigation cycle, NO3 N could have been lost through denitrification or leaching. The increased accumulation of NO3 N in rice 2008 at harvest in all treatments could be due to no further irrigation around maturity; this was not observed in 2009, as the field was intermittently irrigated up to the rice harvest to allow germination of the wheat. The higher N output than N input during the wheat growing season suggests no loss of N from the wheat field. The exhausted soil fertility during the wheat growing period, i.e., reduction of soil total N at wheat harvest compared to seeding (0.031 vs. 0.044%, 29%) and available phosphorus (18.2 vs. 22.4 mg kg−1 , 19%) may explain the higher N output than input during the growing season. Some N could also have been added from residue decomposition and irrigation water, which was not included in the calculation of unaccounted mineral N. Hu et al. (1999) also found higher N output than input during the wheat season due to an 8% reduction of total N in China. As the soil was continuously under aerobic conditions during the wheat growing season (Fig. 9), predominantly NO3 N should have been present in the soil profile. In contrast, the consistently lower NO3 N compared to NH4 N in the top 80 cm soil profile could be due to the fact that NO3 N is the preferred form of N for uptake by wheat (Hamid, 1972). Single or combination of the following could have accelerated NO3 N uptake by the crop neither the loss and possibly justifies the consistently lower NO3 N content in the soil: (1) The higher plant population (750 plants m−2 ) and more vigorous crop growth with more than 7.0 t ha−1 grain yield; (2) N uptake of more than 145 kg ha−1 (Foulkes et al., 2009); (3) no moisture stress during growing period (Asseng et al., 2004); (4) average temperature between 16 and 28 ◦ C during March–May (main period of fertilizer application) (Porter and Semenov, 2005); (5) deep wheat roots (up to 1.54 m) (Palta et al., 2007); and (6) no unaccounted mineral N during the growing period (Devkota, 2011) possibly justifies the consistently lower NO3 N content in the soil due to uptake by the crop neither the loss. The consistently high NH4 N than NO3 N during both rice and wheat growing season could be attributed to the NH4 N from the irrigation water and groundwater. The 4–6 mg l−1 NH4 N in irrigation water, more than 5 mg l−1 NH4 N in groundwater, shallow groundwater table (1.12 m during rice 2008, 1.84 m during wheat 2009 and 0.94 m during rice 2009) and the frequent irrigation in both crops may explain for this. Devkota (2011b) also reported consistently high NH4 N in cotton–wheat cropping system. 4.2. Effect of establishment method and residue level Nitrogen depletion from rice field has been aggravated by high level of residue retention. In the rice season, the higher unaccounted mineral N in the residue-retained treatments could be related to the reduced plant N uptake as a result of reduced growth (Devkota, 2011) and low soil mineral N availability (Fig. 2B). This has previously been associated with the decreased soil temperature under residue retention (Dusserrea et al., 2012), which results in reduced growth and development (Devkota, 2011). The high amount of residue retention could have immobilized N (Quemada and Cabrera, 1995) or higher amount of N could have lost through N2 O denitrification (Granli and Bockman, 1994), and also through NH3 volatilization (Rice and Smith, 1982; Patra et al., 2004; Rochette et al., 2009). Cumulative NH3 volatilization losses are also greater in no-till with crop residue than in conventional as urease activity is high (Al-Fiananir and Maclftnzie, 1992; Rochette et al., 2009; Dusserrea et al., 2012). The higher amount of unaccounted mineral N in rice 2009 compared to 2008 in the R100 and R50 treatments could be related to

K.P. Devkota et al. / Europ. J. Agronomy 47 (2013) 65–76

loss of N through various pathways. At the onset of the experiment in 2008, obviously no crop residues from the previous crop were available. Therefore, chopped wheat residues were imported and uniformly spread over the plot soil surface in the R50 and R100 treatments at 1.5 t ha−1 and 3.0 t ha−1 , respectively. This amount was not sufficient to cover the soil surface. In 2009, standing residues from rice 2008 and wheat 2009 were retained; at the time of sowing the 2nd rice crop, the cumulative crop residues amounted to 8.5 t ha−1 in the R50 and 14.3 t ha−1 in the R100 treatment were retained (Table 3). Rice and wheat residue can immobilize up to 34% of the applied N (Quemada and Cabrera, 1995), and the N rate should be increased by 15% under residue-retained zero-till directseeded conditions (Ladha et al., 2009a). A residual effect of residue retained DSR treatments of rice on soil mineral N content in wheat indicates release of some amount of N by mineralization (Tonitto et al., 2006). Further, the narrower difference of unaccounted mineral N between DSR and WSR rice in 2008 (Fig. 2) could possibly be due to drilling of N at planting in the DSR treatments. NH3 emissions are generally negligible from direct-seeded rice where majority of fertilizer is incorporated into the soil prior to flooding (Humphreys et al., 1988). In contrast, in 2009, N was not drilled (Table 3) as the field was too dry and hard to drill, and all N was broadcast applied in 4 splits after rice emergence. Split application of N fertilizer increases N use efficiency in rice (Reddy and Patrick, 1976; Cassman et al., 1998). However, under the presence of crop residues at the surface of no-till soils also decrease contact of the urea granules with the soil and also possibly reduce adsorption of NH4 N on soil particles (Rochette et al., 2009) and this may further increase volatilization loss of N (Palma et al., 1998). Thus, it indicates under no-till condition with residue retention, possible N loss could be less under direct drilling of N fertilizer than even under broadcast split application. Intensive soil tillage accelerates N mineralization of crop residues and soil organic N (Sainju and Singh, 2001) and increases accumulation of NO3 N in the soil profile (Al-Kaisi and Licht, 2004). However, the non-significant difference in amount of unaccounted mineral N (Figs. 2–5), and NO3 and NH4 N (Fig. 7) content among the WSRF-R0-AWD, DSRB-R0-AWD and DSRF-R0-AWD treatments during the rice growing season in 2009 indicates that availability of soil mineral N is not affected by establishment method. 5. Summary and Conclusions In irrigated drylands, changing from conventional water seeded rice (WSR) to conservation agriculture practices such as direct seeded rice (DSR) can have a major impact on N dynamics in the soil and on plant N uptake by rice. The N loss in rice is higher under DSR with AWD irrigation and residue retention followed by without residue retention. N loss was lowest under WSR. Denitrification leaching, volatilization and N immobilization are likely to be the major pathways of N loss in DSR with residue retention. As major amounts of N loss was before flowering stage of rice. In conservation based rice–wheat system, majority of N loss could be from rice field while no or very less N loss from wheat field. The low level of soil NO3 N during the wheat growing season was due to uptake by the crop neither the loss. Strategies need to be developed for appropriate N fertilizer management, irrigation scheduling and residue management to increase mineral N availability in DSR with residue retention before these practices can be recommended. Acknowledgements The German Ministry for Education and Research (BMBF) funded this study. This paper includes research results made possible by

75

the ZEF/UNESCO project entitled: Economic and Ecological Restructuring of Land- and Water Use in Khorezm Region (Uzbekistan): A Pilot Project in Development Research. Comments and suggestions from two anonymous reviewers substantially improved the clarity of this manuscript. References Al-Fiananir, T., Maclftnzie, A.F., 1992. Effect of tillage practices and hay straw on ammonia volatilization from nitrogen fertilizer solutions. Canadian Journal of Soil Science 72, 145–157. Al-Kaisi, M., Licht, M.A., 2004. Effect of strip tillage on corn nitrogen uptake and residual soil nitrate accumulation compared with no-tillage and chisel plow. Agronomy Journal 96, 1164–1171. Asseng, S., Jamieson, P.D., Kimball, B., Pinter, P., Sayre, K., Bowden, J.W., Howden, S.M., 2004. Simulated wheat growth affected by rising temperature, increased water deficit and elevated atmospheric CO2 . Field Crops Research 85, 85–102. Aulakh, M.S., Khera, T.S., Doran, J.W., Bronson, K.F., 2001. Denitrification, N2 O and CO2 fluxes in rice–wheat cropping system as affected by crop residues, fertilizer N and legume green manure. Biology and Fertility of Soils 34, 375–389. Bateman, E.J., Baggs, E.M., 2005. Contributions of nitrification and denitrification to N2 O emissions from soils at different water-filled pore space. Biology and Fertility of Soils 41, 379–388. Belder, P., Bouman, B.A.M., Spiertz, J.H.J., Peng, S., Castaneda, A.R., Visperas, R.M., 2005. Crop performance, nitrogen and water use in flooded and aerobic rice. Plant and Soil 273, 167–182. Bergstrom, L., Johansson, R., 1991. Leaching of nitrate from monolith lysimeters of different types of agricultural soils. Journal of Environmental Quality 20, 801–807. Beyrouty, C.A., Grigg, B.C., Norman, R.J., Wells, B.R., 1994. Nutrient uptake by rice in response to water management. Journal of Plant Nutrition 17 (1), 39–55. Bouwman, A.F., 1996. Direct emission of nitrous oxide from agricultural soils. Nutrient Cycling in Agro-ecosystem 46, 53–70. Bremner, J.M., Mulvaney, C.S., 1982. Nitrogen-total. In: Page, A.L. (Ed.), Methods of Soil Analysis: Part 2. Chemical and Microbiological Properties. Agronomy Monograph Number 9, ASA and SSSA, Madison, WI, pp. 595–624. Buresh, R.J., De Datta, S.K., 1991. Nitrogen dynamics and management in rice–legume cropping systems. Advances in Agronomy 85, 316–321. Buresh, R.J., Haefele, S.M., 2010. Changes in paddy soils under transition to watersaving and diversified cropping systems. In: 19th World Congress of Soil Science, Soil Solutions for Changing World, 1–6 August 2010, Brisbane, Australia. Buresh, R.J., Woodhead, T., Shephard, K.D., Flordelis, E.V., Cabangon, R.C., 1989. Nitrate accumulation and loss in mungbean/lowland rice cropping system. Soil Science Society of America Journal 53, 477–482. Cassman, K.G., Peng, S., Olk, D.C., Ladha, J.K., Reichardt, W., Dobermann, A., Singh, U., 1998. Opportunities for increased nitrogen-use efficiency from improved resource management in irrigated rice systems. Field Crops Research 56, 7–39. Christmann, S., Martius, C., Bedoshvili, D., Bobojonov, I., Carli, C., Devkota, K., Ibragimov, Z., Khalikulov, Z., Kienzler, K., Manthrithilake, H., Mavlyanova, R., Mirzabaev, A., Nishanov, N., Sharma, R.C., Tashpulatova, B., Toderich, K., Turdieva, M., 2009. Food Security and Climate Change in Central Asia and the Caucasus. CGIAR-PFU, ICARDA, Tashkent, Uzbekistan, p. 75. Devkota, K.P., 2011. Resource Utilization and Sustainability of Conservation Based Rice–Wheat Cropping Systems in Central Asia. ZEF/Rheinische FriedrichWilhelms-Universität Bonn, Germany, p. 167. Devkota, M.D., 2011b. Nitrogen Management in Irrigated Cotton-Based Systems Under Conservation Agriculture on Salt-Affected Lands of Uzbekistan. ZEF/Rheinische Friedrich-Wilhelms-Universität Bonn, Germany, p. 167. Dusserrea, J., Choparta, J.L., Douzeta, J.M., Rakotoarisoab, J., Scopela, E., 2012. Upland rice production under conservation agriculture cropping systems in cold conditions of tropical highlands. Field Crops Research 138, 33–41. FAOSTAT, 2010. FAO Statistics Division. http://faostat.fao.org/site/567/Desktop Default.aspx?PageID=567#ancor Farooq, M., Siddique, K.H.M., Rehman, H., Aziz, T., Lee, D.-J., Wahid, A., 2011. Rice direct seeding: experiences, challenges and opportunities. Soil and Tillage Research 111, 87–98. Firestone, M.K., Davidson, E.A., 1989. Microbiological basis of NO and N2 O production and consumption in soil. In: Andreae, M.O., Schimel, D.S. (Eds.), Exchange of Trace Gases between Terrestrial Ecosystems and the Atmosphere. Wiley and Sons Ltd., Chichester, pp. 7–21. Foulkes, M.J., Hawkesford, M.J., Barraclough, P.B., Holdsworth, M.J., Kerr, S., Kightley, S., Shewry, P.R., 2009. Identifying traits to improve the nitrogen economy of wheat: recent advances and future prospects. Field Crops Research 114, 329–342. George, T., Ladha, J.K., Buresh, R.J., Garrity, D.P., 1992. Managing native and legume fixed nitrogen in lowland rice-based cropping systems. Plant and Soil 141, 69–91. George, T., Ladha, J.K., Garrity, D.P., Buresh, R.J., 1993. Nitrate dynamics during aerobic soil phase in lowland rice-based cropping systems. Soil Science Society of America Journal 57, 1526–1532. Granli, T., Bockman, O.C., 1994. Nitrous oxide from agriculture. Norwegian Journal Agricultural Science, 12. Gupta, R., Kienzler, K., Martius, C., Mirzabaev, A., Oweis, T., de Pauw, E., Qadir, M., Shideed, K., Sommer, R., Thomas, R., Sayre, K., Carli, C., Saparov, A., Bekenov, M.,

76

K.P. Devkota et al. / Europ. J. Agronomy 47 (2013) 65–76

Sanginov, S., Nepesov, M., Ikramov, R., 2009. Research Prospectus: A Vision for Sustainable Land Management Research in Central Asia. CGIAR-PFU, Tashkent, Uzbekistan. Hamid, A., 1972. Efficiency of N uptake by wheat, as affected by time and rate of application, using 15 N-labelled ammonium sulphate and sodium nitrate. Plant and Soil 37, 389–394. Hu, F., Yang, M., Liang, Y., Liu, M., Chen, X., 1999. A study on fertility characteristics of paddy soils under plastic film mulching condition. In: Soil Committee of Jiangsu Province (Ed.), Soil Science Towards 21st Century (Jiangsu Vol.). Hehai University Press, Nanjing, Abstract in English. Humphreys, E., Freney, J.R., Muirhead, W.A., Denmead, O.T., Simpson, J.R., Leuning, R., Trevitt, A.C.F., Obcemea, W.N., Wetselaar, R., Cai, G.-X., 1988. Loss of ammonia after application of urea at different times to dry-seeded, irrigated rice. Fertilizer Research 16, 47–57. Jat, M.L., Gathala, M.K., Ladha, J.K., Saharawat, Y.S., Jat, A.S., Kumar, V., Sharma, S.K., Gupta, R., 2009. Evaluation of precision land leveling and double zero-till systems in the rice–wheat rotation: water use, productivity, profitability and soil physical properties. Soil & Tillage Research 105, 112–121. Johnson-Beebout, S.E., Angeles, O.R., Alberto, M.C.R., Buresh, R.J., 2009. Simultaneous minimization of nitrous oxide and methane emission from rice paddy soils is improbable due to redox potential changes with depth in a greenhouse experiment without plants. Geoderma 149, 45–53. Keeney, D.R., Follett, R.F., 1991. Overview and introduction. In: RFF (Ed.), Managing Nitrogen for Groundwater Quality and Farm Profitability. SSSA, Madison, WI, pp. 1–7. Kyaw, K.M., Toyota, K., Okazaki, M., Motobayashi, T., Tanaka, H., 2005. Nitrogen balance in a paddy field planted with whole crop rice (Oryza sativa cv, Kusahonami) during two rice-growing seasons. Biology and Fertility of Soils 42, 72–82. Ladha, J.K., Kumar, V., Alam, M., Sharma, S., Gathala, M., Chandna, P., Saharawat, Y.S., Balasubramanian, V., 2009a. Integrating crop and resource management technologies for enhanced productivity, profitability, and sustainability of the rice–wheat system in South Asia. In: Ladha, J.K. (Ed.), Integrated Crop and Resource Management in the Rice–Wheat System of South Asia. IRRI, Los Banos, Philippines, pp. 69–108. Ladha, J.K., Kumar, V., Alam, M.M., Sharma, S., Gathala, M.K., Chandna, P., Saharawat, Y.S., Balasubramanian, V., 2009b. Integrating crop and resource management for enhanced productivity, profitability, and sustainability of the rice–wheat system in South Asia. In: Ladha, J.K., Yadvinder-Sing, Erenstein, O., Hardy, B. (Eds.), Integrated Crop and Resource Management in the Rice–Wheat System of South Asia. International Rice Research Institute, Los Banos, Philippines, pp. 69–108. Linn, D.M., Doran, J.W., 1984. Effect of water-filled pore space on carbon dioxide and nitrous oxide production in tilled and no-tilled soils. Soil Science Society of America Journal 48, 1267–1272. Liu, X.J., Wang, J.C., Lu, S.H., Zhang, F.S., Zeng, X.Z., Ai, Y.W., Peng, S.B., Christie, P., 2003. Effects of non-flooded mulching cultivation on crop yield, nutrient uptake and nutrient balance in rice–wheat cropping systems. Field Crops Research 83, 297–311. Merck, 2013. Simply rapid: Merckoquant – the mobile analysis system. Merck KGaA 64271 Darmstadt, Germany. p. 20. Available from http://www.produktinfo. conrad.com/datenblaetter/100000-124999/107271-in-01-en-Nitrat Teststaebchen.pdf (last accessed 23.01.13). Mielke, L.N., Doran, J.W., Richards, K.A., 1986. Physical environment near the surface of plowed and no-tilled soils. Soil and Tillage Research 7, 355–366. Ntanos, D., 2001. Strategies for rice production and research in Greece. In: Chatigner, J. (Ed.), Research Strategies for Rice Development in Transition Economies, CIHEAM-IAMM. Montpellier, France, pp. 115–122. Palma, R.M., Saubidet, M.I., Rìmolo, M., Utsumi, J., 1998. Nitrogen losses by volatilization in a corn crop with two tillage systems in the Argentine pampa. Commununication in Soil Science and Plant Analysis 29, 2865–2879. Palta, J.A., Fillery, I.R.P., Rebetzke, G.J., 2007. Restricted-tillering wheat does not lead to greater investment in roots and early nitrogen uptake. Field Crops Research 104, 52–59. Pandey, S., Velasco, L., 1999. Economics of direct seedling in Asia, Patterns of adoption and research priorities. International Rice Research Newsletter 24, 6–11. Pandey, S., Velasco, L., 2002. Economics of direct seeding in Asia: patterns of adoption and research priorities. In: Pandey, S., Mortimer, M., Wade, L., Tuong, T.P.,

Lope, K., Hardy, B. (Eds.), Direct Seeding: Research Strategies and Opportunities. International Rice Research Institute, Los Banos, Philippines. Pathak, H., Li, C., Wassmann, R., Ladha, J.K., 2006. Simulation of nitrogen balance in rice–wheat systems of the Indo-Gangetic plains. Soil Science Society of America Journal 70, 1612–1622. Patra, A.K., Chhonkar, P.K., Khan, M.A., 2004. Nitrogen loss and wheat (Triticum aestivum L.) yields in response to zero-tillage and sowing time in a semiarid tropical environment. Journal of Agronomy and Crop Science 190, 324–331. Patrick, W.H., Wyatt, R., 1964. Soil nitrogen loss as a result of alternate submergence and drying. Soil Science Society of America Journal 28, 647–653. Pittelkowa, C.M., Fischera, A.J., Moechniga, M.J., Hill, J.E., Kofflera, K.B., Muttersb, R.G., Greerb, C.A., Choa, Y.S., van Kessela, C., Linquista, B.A., 2012. Agronomic productivity and nitrogen requirements of alternative tillage and crop establishment systems for improved weed control in direct-seeded rice. Field Crops Research 130, 128–137. Porter, J.R., Semenov, M.A., 2005. Crop responses to climatic variation. Philosophical Transaction Royal Society of B 360, 2021–2035. Protasov, P.V., 1977. Methods of Agro-Chemical Analyses of Soil and Plants. All Union Research Institute of Cotton Science, Tashkent. Quemada, M., Cabrera, M.L., 1995. Carbon and nitrogen mineralized from leaves and stems of four cover crops. Soil Science Society of America Journal 59, 471–477. Reddy, K.R., Patrick, W.H., 1976. Effect of frequent changes in aerobic and anaerobic conditions on redox potential and nitrogen loss in a flooded soil. Soil Biology and Biochemistry 8, 491–495. Rice, C.W., Smith, M.S., 1982. Denitrification in no-till and plowed soils. Soil Science Society of America Journal 46, 1168–1173. Rochette, P., Angers, D.A., Chantigny, M.H., MacDonald, J.D., Bissonnette, N., Bertrand, N., 2009. Ammonia volatilization following surface application of urea to tilled and no-till soils: a laboratory comparison. Soil and Tillage Research 103, 310–315. Sadanandan, N., Mahapatra, I.C., 1973. Studies in multiple cropping-balance sheet of nitrogen in various cropping patterns. Indian Journal of Agronomy 18, 323–327. Saharawat, Y.S., Singh, B., Malik, R.K., Ladha, J.K., Gathala, M., Jat, M.L., Kumar, V., 2010. Evaluation of alternative tillage and crop establishment methods in a rice–wheat rotation in North Western IGP. Field Crops Research 116, 260–267. Sainju, U.M., Singh, B.P., 2001. Tillage, cover crop, and kill-planting date effects on corn yield and soil nitrogen. Agronomy Journal 93, 878–886. SAS-Institute, 2008. SAS 9.2 Copyright 2002–2008. SAS Institute Inc., Cary, NC, USA. Scheer, C., Wassmann, R., Butterbach-Bahl, K., Lamers, J.P.A., Martius, C., 2008. The relationship between N2 O, NO, and N2 fluxes from fertilized and irrigated dryland soils of the Aral Sea basin, Uzbekistan. Plant and Soil 314 (1–2), 273–283. Scholefield, D., Hawkins, J.M.B., Jackson, S.M., 1997. Use of a flowing helium atmosphere incubation technique to measure the effects of denitrification controls applied to intact cores of a clay soil. Soil Biology and Biochemistry 29, 1337–1344. Sudhir-Yadav, Gill, G., Humphreys, E., Kukal, S.S., Walia, U.S., 2011a. Effect of water management on dry seeded and puddled transplanted rice, part 1: crop performance. Field Crops Research 120, 112–122. Sudhir-Yadav, Humphreys, E., Kukal, S.S., Gill, G., Rangarajan, R., 2011b. Effect of water management on dry seeded and puddled transplanted rice: part 2: water balance and water productivity. Field Crops Research 120, 123–132. Timsina, J., Connor, D.J., 2001. Productivity and management of rice–wheat systems: issues and challenges. Field Crops Research 69, 93–132. Timsina, J., Singh, U., Badaruddin, M., Meisner, C., Amin, M.R., 2001. Cultivar, nitrogen, and water effects on productivity, and nitrogen-use efficiency and balance for rice–wheat sequences of Bangladesh. Field Crops Research 72, 143–161. Tonitto, C., David, M.B., Drinkwater, L.E., 2006. Replacing bare fallows with cover crops in fertilizer-intensive cropping systems: a meta-analysis of crop yield and N dynamics. Agriculture, Ecosystems and Environment 112, 58–72. Vlek, P.L.G., Byrnes, B.H., 1986. The efficacy and loss of fertilizer N in lowland rice. Fertilizer Research 9, 131–147. Wludyka, P., 2011. Study designs and their outcomes. In: Macha, K., McDonough, J.P. (Eds.), Epidemiology for Advanced Nursing Practice. Jones & Bartlett Learning, USA, pp. 81–115.