Nitrogen budgets for Boro rice (Oryza sativa L.) fields in Bangladesh

Field Crops Research 131 (2012) 97–109

Contents lists available at SciVerse ScienceDirect

Field Crops Research journal homepage: www.elsevier.com/locate/fcr

Nitrogen budgets for Boro rice (Oryza sativa L.) fields in Bangladesh M.F. Hossain a,b,c,∗ , S.F. Elahi d , S.K. White c , Q.K. Alam e , J.A. Rother f , J.L. Gaunt c a

American International University-Bangladesh, House 58/B, Road 21, Kemal Ataturk Avenue, Banani, Dhaka 1213, Bangladesh Natural Resources Canada, ESS/CCRS/EMS/AD, 588 Booth Street, Room 423, Ottawa, K1A 0Y7 ON, Canada Rothamsted-Research, Soil Science Department, Harpenden, Herts, UK d Department of Soil, Water and Environment, University of Dhaka, Dhaka 1000, Bangladesh e PROSHIKA, I/1 GA, Section 2, Mirpur 2, Dhaka 1216, Bangladesh f NRI, Chatham Maritime, Kent, UK b c

a r t i c l e

i n f o

Article history: Received 14 December 2011 Received in revised form 16 February 2012 Accepted 17 February 2012 Keywords: Nitrogen budgets Rice Lowland systems Ecological and conventional farming Bangladesh

a b s t r a c t Nitrogen (N) budgets are a valuable tool for improving N efficiency because they assess the size and interactions of various N pools, as well as their gains from the atmosphere and losses to the environment. To understand the impact of changes in management practice upon a farming system, it is necessary to increase the complexity of the N budgets to include N flows. Therefore, a project was undertaken in lowland irrigated systems of Bangladesh to study the N budgets of Boro rice grown under ecological and conventional farming systems in four locations (Dhamrai, Daulatpur, Gabtali and Shibgonj) in Bangladesh in 2007 and 2008. The N budget focuses on the total-N inputs and losses of the entire system. The budgets were negative for both farming systems in both years. Overall, ecological farming system produced a less negative balance in both years (−6 to −36 kg N ha−1 in 2007 and −76 to −160 kg N ha−1 in 2008) than the conventional farming system (−28 to −80 kg N ha−1 in 2007 and −91 to −157 kg N ha−1 in 2008). Nitrogen balance studies highlighted losses of mineral N (26–53 kg N ha−1 ) which accumulated prior to irrigation and also losses due to N removal (13–28 kg N ha−1 ) by weeds. Beneficial impacts of ecological farming on N balances were observed due to the elimination of fertiliser N loss (30–133 kg N ha−1 ). The difference between conventional and ecological management reflects the high losses of fertiliser N under conventional management. These fertiliser N losses reflect the low agronomic efficiency of N fertiliser. An understanding of various N losses and their consequences is important to provide a basis for developing efficient N management strategies in boro rice. These N budgets can be used to improve or design new technologies that tackle soil fertility management problems and also can help improve the financial performance of the farmers. Soil N budgets will continue to challenge agricultural scientists by slowly revealing fundamental principles. By understanding these principles and the factors influencing them, basic and applied scientists will have a stronger foundation for improving N use efficiency and concurrently reducing N losses to the environment. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Nutrient budget studies may differ in purpose, budgeting approach, scale and data acquisition strategy (Song et al., 2011; Ma et al., 2010; Smaling and Oenema, 1997; Smaling, 1998; Regmi et al., 2002; Smaling and Fresco, 1993; Ladha et al., 2003; Kyaw et al., 2005). The purpose of the study defines the approach, scale,

∗ Corresponding author at: American International University-Bangladesh, House 58/B, Road 21, Kemal Ataturk Avenue, Banani, Dhaka 1213, Bangladesh. Tel.: +88 02 8820865/9890804x140; fax: +88 02 8813233; mobile: +88 01759316682. E-mail addresses: [email protected], [email protected] (M.F. Hossain). 0378-4290/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.fcr.2012.02.029

and data acquisition strategies. Large scale budgets for example, at ‘supra-national’ (Stoorvogel et al., 1993), national (Slak et al., 1998) and district scale (Smaling et al., 1993) can help in making policy decisions and often emphasize the need for integrated nutrient management systems. Nutrient balances require inputs and outputs related to the system to be quantified and are the simplest method of compiling information on nutrients in different management systems (Ghaley et al., 2010; Zhao et al., 2009; Langeveld and Overbosch, 1996; Timsina et al., 2006a; Regmi et al., 2002; Kyaw et al., 2005). The balance of inputs and outputs to the system can be a measure of sustainability or efficiency of the management practice. However, farm balances do not give information for internal processes, for example, soil biological processes and can lead to a failure in understanding the changes certain management practices would introduce, such as the use of legumes. Therefore budget

98

M.F. Hossain et al. / Field Crops Research 131 (2012) 97–109

complexity can be increased, for example, to whole field budgets with internal and external nutrient fluxes. Determining whether nutrient inputs are of adequate amount to compensate for outputs, can be used as a measure of the sustainability of a farming system and is gaining importance internationally as an indicator of sustainable land management (FAO, 2003). In Asia, three decades of intensive cropping following the ‘Green Revolution’ has focused Governments and developing agencies on the goal of increasing crop yields of improved varieties using regular applications of fertiliser at recommended rates (Price and Balasubramanian, 1998). Crop yields measured on research stations are rarely achieved in farmers’ fields, because farmer knowledge or understanding has not kept pace with the technology. Consequently nutrient inputs for both the crop and soil are poorly managed as farmers have difficulty in identifying or understanding the cause. In addition soil fertility is declining due to intensive cultivation of rice crops (Gami et al., 2001; Hossain, 2001; Hossain et al., 2003; Ladha et al., 2003; Timsina et al., 2006b). Farmers rely on chemical fertilisers to grow high yielding varieties and use sources of organic fertiliser for other means such as animal bedding and fuel. In Bangladesh, the non-government organization PROSHIKA, have been promoting ‘ecological farming’ throughout the country in an attempt to re-educate people in the benefits of adding organic matter to the soil. The basis of ecological farming is recycling of organic materials found inside the homestead boundary back to the soil to improve or maintain soil fertility. Households apply only organic matter in the form of compost, farm yard manure (FYM) and mustard oilseed cake to fields, no chemical fertilisers or pesticides are used. In general, most organic matter will be applied to the Boro rice crop, grown from mid-February to May, which is the most important high yielding variety of rice. Farmers do not practice ecological farming in isolation, as organic resources are scarce, but will also practice conventional farming. Farmers also incorporate some of the ecological techniques that they find beneficial into the conventional management. The author’s previous study indicates that poor agronomic efficiency of fertiliser N is a limiting factor for crop production in Bangladeshi farmers’ fields (Hossain et al., 2005). Therefore, the objective of this study was to examine further the magnitude of N losses from farmers’ fields under contrasting conventional and ecological farming systems. The study was conducted on Boro rice fields, because this high yielding crop receives the greatest inputs of organic matter and inorganic fertilisers.

2. Materials and methods 2.1. Site selection and characteristics The four sites, selected on the basis of land types and associated flooding regime, had poorly drained soil. The land type was used as selection criterion because of its influence on cropping pattern and farming system. The water management practice, land type and cropping pattern are shown in Table 1. Sites were located in Dhaka (Dhamrai and Daulatpur Thana) and Bogra (Gabtali and Shibganj Thana) greater districts. Field sites were located in thanas (smallest administrative unit), where PROSHIKA (a non-governmental organization) had established Area Development Centers. At each site a local member of PROSHIKA, called an Economic Development Worker (EDW) assisted in maintaining liaison with farmers and monitoring farming practices used in the field and soil. They also collected field data, fertiliser and manure samples. This helped to give comprehensive information of how farming was practiced. Experimental fields were established within farmers’ fields with the following two different farming systems (conventional and

ecological farming) in two different years (2007 and 2008) for Boro rice growing seasons (Table 1):

(i) Conventional farming: the farmers’ current practice using both organic matter inputs and agro-chemicals. In all research areas farmers applied N fertiliser through urea in three equal splits in conventional fields only. The first N application was at final land preparation or 1–2 days after seedling establishment or incorporated into the soil by puddling (tillering and harrowing) or broadcast onto the surface after transplanting. The second application was broadcast at rapid tillering and incorporated in soil along with weeds manually. The third application was broadcast 5–7 days before panicle initiation. P and K were also applied once at the time of final land preparation. (ii) Ecological farming as promoted by PROSHIKA is based on the use of ‘quick compost’ (a mixture of cow dung, rice bran and oil cake in the ratio 4:2:1) and recycling of plant residues to soils instead of reliance on chemical fertilisers and pesticides.

Conventional farming was selected next to the existing ecological farming fields to create pairs on the same type of soil and reduce site variability for samples taken within each site. Historically, both farming systems had natural vegetation on them up to the time of implementation of the experimental sites, after which all treatments have had similar crop rotations throughout the period of study. Hence, it is assumed that any changes over time after the establishment of the experiments can be attributed to differences in land management. The amount of urea N fertiliser applied to the fields was calculated by EDW based on measurements of soil fertility. The fertility status was assessed on the basis of critical limits for soil nutrients based on Bangladesh Agricultural Research Council (BARC) fertiliser recommendation guide (BARC, 2005). Soil data for the conventional and ecological farming fields and the BARC nutrient critical levels used for fertiliser recommendations and nutrient status are shown in Table 2. Experimental details have been reported elsewhere (Hossain et al., 2005). The ecological farming at Dhamrai and Daulatpur had a significantly higher amount of CEC compared to the conventional farming. Ecological farmed managed fields had a greater amount of organic carbon as compared to the conventional farming at Dhamrai and Daulatpur areas. Additional benefits of ecological farming reported by farmers were increased ease of tillage and greater number of arthropods. Incorporating organic matter regularly into fields helps maintain soil fertility, using new technologies is essential to increase yields. Unfortunately, there is a general lack of organic matter available for crop production. Resources being limited to the extent that farmer will limit the number of fields subjected to ecological management. Returning therefore to conventional farming is not an option, enhancing current practices with new technologies is.

2.2. Field sampling and measurements The field sizes of the conventional and ecological farming varied but were not less than 12 m × 12 m areas. At each site, up to 10 replicates of farmer’s fields were randomly selected for both conventional and ecological farming because most of the time, the same farmer managed a conventional and ecological farming field. Each field was surrounded by an approximately 15 cm height bund (‘ieal’, in Bengali) to minimize movement of applied fertiliser from the surrounding fields. Some fields were eliminated from the study because farmers mistakenly applied N fertiliser to the ecological fields. This gave unequal replications between the areas.

M.F. Hossain et al. / Field Crops Research 131 (2012) 97–109

99

Table 1 Site characteristics including land type, irrigation system, land use pattern, and Boro rice growing season (seedling transplanting and harvesting time). District

Sites

Land type

Irrigation system

Land use pattern

Transplanting and harvesting time

No. of fields

Dhaka

Dhamrai

MLL

Irrigated

Boro (BR3) – Fallow

5

LL

Irrigated

Boro (BR3) – Fallow

MLL

Irrigated

Boro (IR8) – Fallow

VLL

Non-irrigated

Boro (IR8) – Fallow

MHL

Irrigated

RC – Boro (BR14) – Jute

Mid December to Mid January and Mid April to Mid May Mid December to Mid January and Mid April to Mid May Mid December to Mid January and Mid April to Mid May Mid December to Mid January and Mid April to Mid May Early January to Early February and late April to Late May

MLL

Irrigated

Boro – B. Aman Rice

HL

Irrigated

RC(Potato) – Boro (BR1) – KC

MHL

Irrigated

RC (Potato) – Jute

Daulatpur

Bogra

Gabtali

Shibganj

5 4 6 5 5

Early January to Early February and late April to Late May

8 2

RC: Rabi (winter) crop for example; lentil, mustard, onion, chilli and potato grown after the monsoon season; KC: kharif crop; HL: high land; MHL: medium high land; MLL: medium low land; LL: low land; VLL: very low land; IR8, BR14, BR3 and BR1: high yielding variety of rice.

2.3. Laboratory analyses

2.4. Statistical analyses

total amount of manure added to each field for all sites were calculated using the mean weight of five FYM baskets. This method of application to the experimental fields at the different research sites was used throughout the experimental period. Four main types of organic manure were used in the experiment for both conventional and ecological fields: farmyard manure, compost, oil cake and quick compost.

The variation in the experimental data of the two different farming systems and two different years was analysed statistically using a two-sample unpaired t-test using Genstat 5 V 4.1 (Lane and Payne, 1998).

2.5.3. Atmospheric deposition of nitrogen Nitrogen input by deposition consists of two parts: dry and wet deposition. Factors of N were calculated based on field measurements.

All chemical analyses (total N: NH4 + -N and NO3 − -N) were determined on the samples using the TSBF colorimetric method (Anderson and Ingram, 1993).

2.5. Inputs 2.5.1. Chemical fertiliser Samples of fertilisers that were to be applied on fields were collected from the research sites. The amount of fertiliser being applied in each field was recorded by participating farmers. Approximately 20 g of urea sample was placed in a labelled polythene bag and brought to the laboratory for chemical analysis. Moisture content was determined on part of the sample by drying at 110 ◦ C. Total N was determined and N inputs estimated by multiplying the kilograms of fertiliser applied per hectare by the N content of the urea. 2.5.2. Organic manure Farmers add farmyard manure, oil cake, and quick compost (QC) using a basket local Bengali name “Tukri” and hence the numbers of baskets of organic manure used in each field were counted. The

2.5.3.1. Dry deposition. Passive diffusion tubes were used to measure the average atmospheric deposition of nitrogen dioxide (Palmes et al., 1976; Atkins et al., 1986) at the four research locations and assuming that all NH3 deposition comes in wet deposition. 2.5.3.2. Wet deposition. Rainfall collectors were placed in an open location well away from obstructions such as building, trees, roads, and animal houses. The bulk collector was constructed according to the design described by Hall (1986). It consisted of a conical plastic funnel of diameter 152 mm which rests on the neck of a 2 l polythene collecting bottle. The upper surface of the collector was 1.75 m above the ground and surrounded by a jacket of polished steel which allowed at least a 15 cm gap. The samples were kept in dark by the jacket.

Table 2 Nutrient status of ecological and conventional farming fields and BARC critical limits. Parameters

BARC critical limit

Nutrient status Dhamrai

TN (%) TOC (%) C:N TP (%) K+ cmol+ kg−1 Ca++ cmol+ kg−1 Mg++ cmol+ kg−1 CEC cmol+ kg−1

0.12 1.20 10:1 0.12 0.12 2.00 0.50 M–H

Daulatpur

Gabtali

Shibganj

Ecological

Conventional

Ecological

Conventional

Ecological

Conventional

Ecological

Conventional

0.20 1.77* 9:1 0.27 0.21 9.30** 2.81 21**

0.18 1.40 8:1 0.23 0.20 6.23 2.31 18

0.17 1.64** 10:1* 0.23 0.21 8.03** 2.65 25*

0.15 1.14 8:1 0.20 0.18 6.41 1.89 21

0.12 1.13 10:1 0.18 0.22 4.61 1.78 13

0.11 1.20 11:1 0.18 0.19 5.49 1.98 12

0.11 1.14 12:1 0.13 0.35 5.22 1.43 14

0.11 1.30 10:1 0.13 0.24 4.10 1.47 14

TN: total nitrogen; TOC: total organic carbon; C:N: carbon nitrogen ratio; TP: total phosphorus; M (medium) = 7.5–15 cmol+ kg−1 ; H (high) = 15–30 cmol+ kg−1 . * Significant different at P < 0.05 in between conventional and ecological fields. ** Significant different at P < 0.01 in between conventional and ecological fields.

100

M.F. Hossain et al. / Field Crops Research 131 (2012) 97–109

The water samples were collected after a heavy rain event and either kept in a refrigerator or acidified with concentrated sulphuric acid. Total N (NH4 + -N and NO3 − -N) was determined. Data obtained from the Bangladesh Meteorological Department (Climate Division, 2007 and 2008) showed that in 2007 and 2008, there was approximately 224 mm and 391 mm of rain for Dhaka region (used for both Dhamrai and Daulatpur sites), respectively, and 115 mm and 219 mm of rain for Bogra region (used for both Gabtali and Sibganj sites), respectively. Deposition of total N in rainfall is calculated by multiplying the concentration of N in the collected water by the volume of rainfall. 2.5.4. Sedimentation This flux consists of two parts: input of nutrients by irrigation water and input of sediments as a result of erosion. The input of N by irrigation water was estimated by field measurements and sedimentation N input was used from literature, which is based on floodplain rice areas especially for Dhamrai and Daulatpur. Gabtali and Shibganj are above the flood levels; therefore, for these two areas sedimentation N inputs are not relevant. In the parts of the naturally flooded areas, sedimentation takes place. Hardly any information on the N content of these sediments could be traced. However, it was necessary to make an assumption on the importance of this input factor. The group of experts reached consensus on N budgets being in equilibrium in land and water classes. Input and output factors were calculated but the deficit sedimentation N input was assumed to be supplied by the flood water and its sediments, and an estimated value of 10 kg N ha−1 year−1 was thus used (FAO, 2003). 2.5.5. Irrigation water Irrigation at the research sites used predominantly shallow tube wells. Water samples were collected for NH4 + -N and NO3 − -N analysis directly from the pumps using plastic bottles (125 ml). The bottle tops were removed just prior to collecting the sample to avoid sample contamination and 2 ml of sulphuric acid were added to prevent microbial activity. The quantity of NH4 + -N and NO3 − -N was calculated by multiplying the amount of water applied to the field by its total N concentrations. 2.5.6. Biological nitrogen fixation (BNF) Bangladesh’s floodplain soils nutrient status states that BGA can provide up to 30 kg N ha−1 year−1 (Brammer, 1983), and perhaps even more on the deeply flooded areas where deep water rice is grown. Plant residues provide an important input of carbon to soils, which can act as a substrate for BNF. For example, following flooding, rice straw addition to soils increased the rate of BNF (Yoneyama et al., 1977; Reddy and Patrick, 1979). Roger and Kulasooriya (1980) estimated the BNF by cyanobacteria to be approximately 27 kg N ha−1 per rice crop with a maximum of 50–80 kg N ha−1 . Such estimates not only reveal a wide range in fixation values, but also reflect the inherent difficulties in measuring BNF under natural conditions. Therefore, the estimates of BNF were obtained from the literature because of logistical difficulties in making such measurements. The contribution of BNF by cyanobacterial communities to the Boro rice crop in Bangladesh near to the research areas of Dhamrai and Daulatpur was estimated to be 10.2 kg N ha−1 during the pre-flood period (Rother and Whitton, 1989). This estimate of cyanobacterial N2 -fixation represents 10 kg N ha−1 for Boro rice growing season in the N budgets. 2.6. Outputs 2.6.1. Crop production and N removal To determine the grain yield and quantity of N removed by the high yielding variety of Boro rice crop samples were taken at final

harvest stage within a 10 m × 10 m area at the centre of the fields, allowing a 1 m edge buffer zone. Within the sampling area, five sampling spots were selected at random using an approximately 2 m2 quadrat area (1.42 m × 1.42 m) in each experimental field. Plants were cut at soil level. Plant samples were threshed immediately after sampling and the total weight of straw and grain per quadrat was recorded. Fresh weights of subsamples were determined to the nearest 0.01 kg in the field using a beam balance. Yields were estimated based on the oven dry (60 ◦ C) weight and adjusted afterwards at a standard moisture content of 14% (Triol Padre et al., 1996). Representative straw and grain subsamples (100–200 g) were collected, oven dried at a maximum temperature of 60 ◦ C to a constant weight, then ground through a 0.15 mm mesh, and stored in a dessicator in labelled paper bags until analysis. Plant samples (grain and straw) were analysed for total N concentrations using micro Kjeldahl digestion techniques. 2.6.2. Weeds Field surveys were conducted to determine the predominant weed species in rice fields. These were identified as Monochoria veginalis, Fimbristylis miliacea, Scirpus spp., Marsilea crenata, Ludwigia octovalvis and Sphenoclea zeylanica. Total weed yield was recorded using a quadrant according to sampling collection and methods described elsewhere (Hossain, 2001). 2.6.3. Estimation of N losses Estimation of N losses due to irrigation may lead to leaching, denitrification and volatilization. We measured only leaching losses for both conventional and ecological farming systems and assumed that gaseous N losses accounted for the remaining. Initially N loss was calculated based on recovery efficiency of N [RE (N)] calculated using the following equation (Cassman et al., 1996; Hossain et al., 2005): %RE (N) =

UN − U0 × 100 Nr

where Nr is the amount of N fertiliser applied, UN is the plant N accumulation (straw and grain N together) with applied N fertiliser, and U0 is the plant N accumulation (straw and grain N together) without N fertiliser applied. Composite soil samples were collected using a PVC tube (length 200 mm, 38 mm i.d., 49 mm o.d.) before land preparation for transplanting. For both management practices, three replicate samples were taken from each field at each site. Fresh moist soil (10 g) was extracted with 0.5 M K2 SO4 after shaking for 30 min immediately after sampling to avoid the accumulation of nitrate due to organic matter mineralisation. The filtrates were stored in a refrigerator or icebox prior to N analysis. Moisture content of the fresh soil was also determined in conjunction with N determinations. 2.6.3.1. Measurement of N leaching by resin bag method. A PVC tube (length 200 mm, 38 mm i.d., 49 mm o.d.) was inserted into the soil for both conventional and ecological farming systems until the top of the tube was levelled with the soil surface (Fig. 1a). A set of four equally spaced holes was drilled around the tube (50 mm from the top and 150 mm from the base of the tube) to provide outlets for excess surface water to drain by lateral flow from the soil. An O-ring was positioned in a grove cut on the outside of the tube immediately below the drainage hole. The metal corer was placed over the top of the protruding incubation tube and driven into the ground to the same depth as the soil core (Fig. 1b) and additional outer core was peeled away carefully from the PVC tube (Fig. 1c). A groove was cut in the top end (Fig. 1d) of the tube (length 175 mm, 50 mm i.d., 60 mm o.d.) and fitted with a gas tight O-ring seal before connecting the two tubes. A cylindrical block (47 mm diameter, 15 mm deep) was placed at the bottom of the vessel. The ion exchange resin

M.F. Hossain et al. / Field Crops Research 131 (2012) 97–109

101

Fig. 1. (a–d) The field incubation system (a) incubation tube a soil core, (b) metal soil cover, (c) baseline and incubation core, and (d) incubation tube fitted into incubation vessel above I-E bag.

(I-E bag) was placed in bags at the base of the core (Fig. 1d). It was sandwiched between two filter paper discs (47 mm diameter, grade 40; Whatman) which protected the resin and improved contact between surfaces. The function of the I-E bag was to intercept any NO3 − N and NH4 + -N which passed through the soil core. The core was incubated in the field with the I-E bag. At the end of each incubation period, the soil core was removed every seven days in the first two months, and then every 15 days up to Boro rice harvest. The NO3 − N and NH4 + -N adsorbed on the I-E bag was extracted by shaking for an hour with 50 ml of 1.5 M H2 SO4 , followed by two rinses with 50 ml batches of the extraction. This procedure was found to consistently recover 90% of the adsorbed NO3 − N and NH4 + -N under controlled laboratory conditions. A factor of 1.11 was used to adjust estimates of leaching, based on those from I-E bags, to take into account NO3 − N and NH4 + -N recovery. Three replicate measurements, each on an individual field in

each site, were made for both conventional and ecological farming systems (Fig. 2). 2.6.3.2. Measurement of soil mineral N and gaseous N losses. In 2008, soil mineral N losses was measured separately as gaseous N and leaching losses to differentiate how quick the amount was leached and gaseous N lost. Buresh et al. (1989) measured soil NO3 − -N present prior to soil flooding and loss by denitrification and volatilisation after flooding within 10–12 days. Most of the fertiliser N not accounted for by uptake after 10 days was likely lost from the system. Such brief residence of applied N in labile pools during peak growth periods makes it difficult to synchronize N supply with crop demand. On the other hand, uptake by the rice appears to be rapid when an adequate N supply is matched with crop demand (Cassman et al., 1993). To avoid overestimating losses, in 2008, gaseous N loss was estimated from NO3 − -N accumulated prior to

102

M.F. Hossain et al. / Field Crops Research 131 (2012) 97–109

Farmer questionnaire

Literature values

Field measurements

Database

N balance tables

Flow diagram

What-if scenarios

Fig. 2. Diagram to show methodology used to collect data to construct nitrogen budgets.

flooding minus measured NO3 − -N leaching during the 12 days after flooding. During this period when crop growth was low the cores are likely to give a reliable estimate of N leaching. In the final N budgets, estimated gaseous N loss and measured amounts of leaching loss during cropping season were included separately. 2.7. Construction of N budgets Nitrogen budgets were calculated as N inputs minus N outputs,  and equal  the change of N within the system (N budgets = Inputs − Outputs). Data were collected from the field samples, in-field measurements and literature derived values. Inputs considered include chemical fertiliser, organic manures, atmospheric depositions (wet and dry), irrigation, sedimentation and biological N fixation. Outputs include N removed by crop (straw and grain) and weeds, and N losses through leaching, denitrification and volatilisation. 3. Results 3.1. Inputs 3.1.1. Chemical fertiliser The N content of urea fertiliser analysed in the sample was comparable to the content given in the fertiliser tag (46%). Urea N inputs applied to the soils by farmers were variable despite the recommendations made by BARC (BARC, 2005). Fertiliser composition was consistent with the manufacturer specification and did not vary between year or site. This finding was important because there are concerns in Bangladesh that adulterated fertilisers are being supplied. Significantly different (P < 0.001) amounts of fertiliser were applied at the Gabtali site between 2007 and 2008. For high rice yields, BARC recommend an N application rate of 100 kg ha−1 , but the N rates used by farmers ranged from 39 to 175 kg ha−1 . It seems farmers did not follow the BARC recommendation, may be due to ignorance and illiteracy. 3.1.2. Organic manure The average N content of FYM ranged from 1.23 to 1.76%, which was comparable to literature values 0.5–1.5% (BARC, 2005). The only exception was at Gabtali in 2008 where the N concentration of FYM was 0.31%. This was significantly lower than at the other sites in 2008 and lower than measured at Gabtali in 2007. This may be because of improper storage. The heavy monsoon rains washed away manure and led to deposition of sediments in FYM pits. Despite the fact that preparation of quick compost (QC) follows a recipe, N concentration in QC ranged from 0.4 to 1.8%. These results

suggest that variation in inputs used (source, type and composition) or management of the composting process can lead to variation in the quality of the QC produced. It is notable that in 2008 the QC prepared by PROSHIKA at Daulatpur had a significantly higher N concentration (1.8%) than compost prepared by farmers at Dhamrai and Shibganj. The N concentration of oilseed cake depends on the species, oil seed maturity and seed quality. A comparison of N concentration showed that the quality did not vary significantly with the average concentration ranging from 5.35 to 5.82%, which is comparable to values in the literature (from 5.1 to 5.2%; BARC, 2005). This suggests that variation in the quality of QC was not due the quality of the oil seed cake used. The most common green manure crops used in Bangladesh are dhaincha (Sesbania cannabina), lentil (Lens culinaris), sunhemp (Crotalaria juncea), chickpea (Cicer arietinum), cowpea (Vigna unguiculata) but the current study used only Sesbania cannabina for both conventional and ecological fields in each site. The measured N concentration in Sesbania cannabina crop (0.68–0.69%) is comparable to literature value (0.62%; BARC, 2005). 3.1.3. Dry deposition Mean NO3 -N concentrations in 2007 ranged from 1.17 to 1.96 ␮g m−3 at Dhamrai, 1.06 to 1.70 ␮g m−3 at Daulatpur, 0.69 to 1.96 ␮g m−3 at Gabtali and 0.74 to 1.32 ␮g m−3 at Shibganj (Table 3). In 2008, NO3 -N concentration increased significantly (P < 0.001) 3.13–6.89 ␮g m−3 and 1.53–7.14 ␮g m−3 at Dhamrai and Daulatpur, respectively (Table 3). Overall, the NO3 -N concentrations were very low when compared to data from IACR-Rothamsted in the South East of UK (15–25 ␮g m−3 ; Campbell, 1988; Goulding et al., 1998; Hargreaves et al., 2000). The higher concentrations reflected the high density of NOx emissions in the more densely populated and industrialised areas of England (Campbell, 1988). The reason for the higher NO3 -N concentrations in the 2008 Boro growing season at Dhamrai and Daulatpur is not clear. It could be that the NO2 is not emitted from the soil but NO, which reacts in the atmosphere and in certain conditions, produces NO2 . Diffusion tubes will also recover NO2 emitted from soil, so differences could be associated with greater losses from soil associated with water-logging and denitrification losses (Vitousek, 1994). Another explanation is the relative proximity of these two sites to Dhaka city and potential inputs from the city depending on variations in wind direction and pollutant source (Campbell, 1988). In contrast, Gabtali and Shibganj exhibited lower NO3 -N than Dhamrai and Daulatpur. At Dhamrai and Daulatpur sites, N inputs were greatest in February and March, particularly in 2008, whereas at the Gabtali and Shibganj sites, inputs were greater in April and May (Table 3). The lands in Gabtali and Shibganj are well drained, in contrast to Dhamrai and Daulatpur (Table 1). During the ripening phase, farmers do not apply irrigation water. Nitrogen deposition is related to soil temperature, with greater amounts of deposition with high soil temperature (Goulding et al., 1998; Perkauskas and Mikelinshiene, 1998). The cumulative dry deposition of N measured from February to May (data not presented) showed that the total N deposition ranged from 0.24 to 0.38 kg ha−1 in 2007 and 0.54 to 0.97 kg ha−1 in 2008. The cumulative dry deposition N curves were similar, with Gabtali and Shibganj showing a marked increase in N inputs in May. However, in 2008, at Dhamrai and Daulatpur N inputs in February were much higher than those in May. 3.1.4. Wet deposition Nitrogen deposition due to very low rainfall during the Boro cropping season appeared to be negligible from an agricultural point of view (Table 4). For example, high N concentrations were

M.F. Hossain et al. / Field Crops Research 131 (2012) 97–109

103

Table 3 Concentrations of NO3 (␮g N m−3 ) as measured by diffusion tubesa in the 2007 and 2008 Boro rice seasons. Sites

Year

February

Dhamrai

2007 2008 2007 2008 2007 2008 2007 2008

1.43 6.89 1.54 7.14 1.96 1.48 1.06 1.32

Daulatpur Gabtali Shibganj a *

± ± ± ± ± ± ± ±

March

0.11 0.55* 0.12 0.43* 0.16 0.14 0.10 0.12

1.17 5.51 1.70 2.61 1.54 1.02 1.01 1.22

± ± ± ± ± ± ± ±

April 0.13 0.34* 0.21 0.23* 0.14 0.11 0.09 0.11

1.96 3.13 1.06 1.57 0.69 2.82 0.74 2.86

May ± ± ± ± ± ± ± ±

0.15 0.36 0.11 0.15 0.09 0.17* 0.08 0.19*

1.59 3.28 1.22 1.53 1.01 3.23 1.32 2.75

± ± ± ± ± ± ± ±

0.14 0.28 0.13 0.15 0.11 0.19* 0.13 0.22*

Mean of 3 diffusion tubes. Significant different at P < 0.001 in between 2007 and 2008.

found in the Bogra (Gabtali and Shibganj) in the pre-monsoon period (Boro cropping duration) due to the influence of dust during a time with low rainfall. In addition, nitrification continues at low rates during part of the dry season and without a plant sink it simply accumulates in the soil. N concentration in rainfall was significantly different in Bogra between 2007 and 2008 (P < 0.01 at Gabtali and Shibganj). The rainfall N concentration ranged from 1.57 to 2.74 ␮g ml−1 in 2007 and 1.70 to 2.63 ␮g ml−1 in 2008 (Table 4). The total N inputs ranged from 3 to 4 kg ha−1 in 2007, but were significantly increased (P < 0.01) to 5–7 kg ha−1 in 2008 in all four sites (Table 4). The greater N inputs in 2008 were due to an increase in total rainfall. Higher N concentrations were found in the rainfall over the period of winter and pre-monsoon months (Mian et al., 1991). However, the wet N deposition ranged from 3 to 7 kg ha−1 in the Boro cropping season, which was comparable to 4.3 kg ha−1 in Bangladesh (Greenland, 1997) but 10 times lower than that at Rothamsted in the UK during 1998 (Goulding et al., 1998). On global scale, dry deposition of N has been estimated to be up to five times greater than wet deposition (Ivens et al., 1988). However, in Bangladesh, dry deposition of N was significantly less than wet deposition. 3.1.5. Irrigation water The N concentrations of pump water show that there was no significant difference between the years. Concentration varied from 0.63 to 0.68 ␮g ml−1 at Dhamrai, 0.59 to 0.60 ␮g ml−1 at Daulatpur, 0.89 to 0.91 ␮g ml−1 at Gabtali and 0.91 ␮g ml−1 at Shibganj sites, and these values are comparable to literature value of 0.92 ␮g ml−1 (Mian et al., 1991). The pump N ranged from 4 to 8 kg ha−1 for conventional and 3–7 kg ha−1 for ecological farming practices, similar to literature value of 5 kg ha−1 (Mian et al., 1991). The relatively small quantity of N added through irrigation did not appear to warrant further efforts to measure the amount of water used in irrigation. The ranges of N input varied from 3 to 8 kg ha−1 with no significant difference between sources. 3.2. Outputs 3.2.1. Nitrogen uptake by crop Nitrogen uptake, grain and straw yield, and N content increased in 2008, and were similar at the Dhamrai and Daulatpur; Gabtali and Shibganj sites. The N uptake is derived from N concentrations of grain and straw. Therefore, grain and straw yields and their N

concentrations are described briefly. In conventional fields, grain yield varied from 4.25 to 5.97 t ha−1 in 2007 and 4.74 to 8.19 t ha−1 in 2008, whilst in ecological fields it ranged from 3.43 to 4.84 t ha−1 in 2007 and from 4.16 to 6.72 t ha−1 in 2008 (Table 5). In conventional fields, straw yield varied from 2.56 to 3.94 t ha−1 in 2007 and from 4.00 to 5.93 t ha−1 in 2008, whilst in ecological fields 1.95 to 3.29 t ha−1 in 1998 and from 4.21 to 6.36 t ha−1 in 2008 (Table 5). In both farming systems, straw yields increased significantly (P < 0.001) in 2008 at Dhamrai and Daulatpur, whereas grain yields decreased significantly at Dhamrai (P < 0.05) but increased significantly at Daulatpur (P < 0.001). Statistically significant differences between conventional and ecological farming systems were found for both grain and straw yields at Dhamrai in 2007 (P < 0.01) and only for straw yields at Daulatpur in 2007 (P < 0.05; Table 5). The N concentrations in grain and straw were generally significantly higher (P < 0.01) in 2008 than in 2007 at all sites except Daulatpur (Table 6), where the concentrations decreased significantly (P < 0.01 and P < 0.001). In Asia, average N concentration is 1.16% in grain and 0.71% in straw (Witt et al., 1999). The measured grain N concentrations in 2007 (1.06–1.18%) were similar to 1.05–1.18% N as reported by Makarim et al. (1994). However, in 2008, N concentration in grain at most sites was markedly higher (0.85–1.46%; Table 6). The N uptake by both grain and straw was higher in all four sites, but not significantly different, with the range varying considerably within each site. In conventional farming grain, N uptake varied from 52 to 75 kg ha−1 , but in ecological farming, it varied from 45 to 78 kg ha−1 . Straw N uptake in conventional farming varied from 19 to 54 kg ha−1 , but in ecological farming, it varied from 14 to 41 kg ha−1 . Together grain and straw represented approximately by 70–122 kg ha−1 of total N removed under conventional farming and 52–110 kg ha−1 of total N removed from ecological farming. These figures are comparable to the recommendation of N for this crop of 108 kg ha−1 (BARC, 2005). Significant differences between conventional and ecological farming were found in grain N uptake at Dhamrai in 2007 (P < 0.001) and straw yields at Dhamrai for both 2007 and 2008 (P < 0.01). The mean grain yield rate increased 2% on average of all four sites but fertiliser N rate increased approximately 25–27%, which is greater than the yield rate. Greater N uptake in 2008 resulted from increased rates of N application but was not reflected in corresponding increases in yield. This suggested that yields were not limited by N (Hossain et al., 2005).

Table 4 The mean N concentrations and inputs by rainfall for conventional and ecological farming fields, 2007 and 2008 Boro rice seasons. Rainfall

Dhamrai 2007

N concentrations (␮g ml−1 ) N inputs (kg N ha−1 ) *

1.57 4

Daulatpur 2008 1.57 6*

Significant different at P < 0.01 in between 2007 and 2008.

2007 1.70 4

Gabtali 2008 1.70 7*

Shibganj

2007

2008

2007

2008

2.74 4

2.63 6*

2.85 3

2.32 5*

104

M.F. Hossain et al. / Field Crops Research 131 (2012) 97–109

Table 5 Mean yield (t ha−1 ), N concentrations (%) and N uptake (kg N ha−1 ) of grain and straw for conventional and ecological fields in 2007 and 2008 Boro rice. Farming systems

Dhamrai 2007

Grain yield 5.97 Conventional 3.43 Ecological Straw yield Conventional 3.33 Ecological 1.95 Grain N concentrations 1.17 Conventional Ecological 1.32 Straw N concentrations Conventional 0.67 Ecological 0.67 Grain N uptake 69 Conventional 40 Ecological Straw N uptake 22 Conventional Ecological 12 a b c * ** ***

± 0.21*** , a ± 0.28

Daulatpur 2008

2007

5.01 ± 0.32 5.33c ± 0.25

4.88 ± 0.29 4.84 ± 0.10

Gabtali 2008

Shibganj

2007

2008

2007

2008

8.19 ± 0.56c 6.72b ± 0.97

4.25 ± 0.35 4.15 ± 0.66

4.96 ± 0.28 5.15 ± 0.25

4.60 ± 0.24 4.16 ± 0.28

4.74 ± 0.24 4.16 ± 0.23

± 0.19** ± 0.33

5.70 ± 0.49c 5.43 ± 0.20c

2.56 ± 0.24* 1.95 ± 0.18

5.93 ± 0.28c 6.36 ± 0.38c

3.55 ± 0.20 3.18 ± 0.45

4.00 ± 0.56 4.76 ± 0.57a

3.94 ± 0.40 3.29 ± 0.21

4.72 ± 0.28 4.21 ± 0.56

± 0.05 ± 0.08*

1.36 ± 0.05b 1.46 ± 0.05

1.11 ± 0.05 1.16 ± 0.02

0.91 ± 0.02b 0.85 ± 0.03c

1.22 ± 0.03 1.23 ± 0.07

1.37 ± 0.06a 1.34 ± 0.03

1.16 ± 0.05 1.09 ± 0.07

1.20 ± 0.03a 1.27 ± 0.08

± 0.03 ± 0.03

0.97 ± 0.07*** 0.58 ± 0.02

0.62 ± 0.06 0.52 ± 0.02

0.58 ± 0.03* 0.50 ± 0.01

0.72 ± 0.08 0.66 ± 0.05

0.92 ± 0.06 0.88 ± 0.06

0.54 ± 0.08 0.61 ± 0.05

0.69 ± 0.03 0.84* ± 0.09

± 3*** ±6

68 ± 5 78 ± 4** , c

54 ± 5 56 ± 2

75 ± 6*** , c 59 ± 11

52 ± 5 51 ± 8

68 ± 5c 69 ± 4c

54 ± 4* 46 ± 5

57 ± 3 53 ± 6c

± 2*** ±2

54 ± 5*** , c 32 ± 2c

16 ± 2*** 10 ± 1

34 ± 2c 32 ± 2c

26 ± 3* 21 ± 3

38 ± 8c 41c ± 5

25 ± 4* 20 ± 2

33 ± 3c 36c ± 7

Significant different in between 2007 and 2008 at P < 0.05. Significant different in between 2007 and 2008 at P < 0.01. Significant different in between 2007 and 2008 at P < 0.001. Significant different at P < 0.05 in between conventional and ecological fields. Significant different at P < 0.01 in between conventional and ecological fields. Significant different at P < 0.001 in between conventional and ecological fields.

Table 6 Boro rice recovery efficiency of applied N [% RE (N)], agronomic efficiency (AE; Hossain et al., 2005) and chemical fertiliser N losses for all fields in 2007 and 2008. Variables

Dhamrai 2007

Recovery efficiency of applied N 21 ± 9 Mean AE ( kg grain kg−1 applied N) Mean 19 ± 5 Chemical fertiliser N losses (kg N ha−1 ) 97 ± 7 Conventional * ** ***

Daulatpur

Gabtali

Shibganj

2008

2007

2008

12 ± 2**

18 ± 3

10 ± 4**

9±3

12 ± 3

7 ± 2**

11 ± 2

5 ± 2**

13 ± 4

83 ± 14

36 ± 3

92 ± 7***

42 ± 2

9 ± 2*** 154 ± 21*

98 ± 14

2007

2008

2007

2008

15 ± 3*

17 ± 4

19 ± 2 9 ± 2* 47 ± 4

Significant different in between 2007 and 2008 at P < 0.05. Significant different in between 2007 and 2008 at P < 0.01. Significant different in between 2007 and 2008 at P < 0.001.

The fact that N content of straw and grain differed significantly under differing management and years suggests that yields were limited by a factor other than N. Thus in this study it would not have been appropriate to calculate N outputs based solely on yield measurements as proposed by Kundu and Ladha (1999). It was also concluded that yields were generally similar in both conventionally and ecologically managed fields with some exceptions.

3.2.2. Nitrogen uptake by weeds The N content of weeds varied considerably between years with significant differences at Shibganj (P < 0.01) in the ecological farming fields. In conventional farming systems it varied from 1.12 to 1.32% in 2007 and from 1.13 to 1.34% in 2008, whilst in ecological farming systems it was 0.91–1.34% in 2007 and 1.31–1.37% in 2008. Significant differences in N concentration were found between conventional and ecological farming fields at

Table 7 The N concentrations (%) and N uptake (kg N ha−1 ) by weed from conventional farming and ecological farming fields in 2007 and 2008 Boro rice seasons. Farming systems

Dhamrai 2007

Weed N concentrations 1.12 Conventional Ecological 1.11 Weed N uptake 24 Conventional 19 Ecological a

± 0.05 ± 0.08 ±3 ±3

Daulatpur

Gabtali

2007

2008

2007

2008

2007

2008

1.13 ± 0.05 1.18 ± 0.05

1.32 ± 0.05 1.28 ± 0.02

1.13 ± 0.02 1.15 ± 0.03

1.28 ± 0.03 1.23 ± 0.07

1.34 ± 0.06* 1.20 ± 0.03

1.12 ± 0.05** 0.91 ± 0.07

1.27 ± 0.03a , ** 0.95 ± 0.08

25 ± 2 23 ± 1

26 ± 1 22 ± 1

27 ± 3 22 ± 1

18 ± 1 22 ± 4

17 ± 1 26 ± 3*

Significant different in between 2007 and 2008 at P < 0.05. Significant different in between 2007 and 2008 at P < 0.01. Significant different in between 2007 and 2008 at P < 0.001. * Significant different at P < 0.05 in between conventional and ecological fields. ** Significant different at P < 0.01 in between conventional and ecological fields. *** Significant different at P < 0.001 in between conventional and ecological fields. b c

Shibganj

2008

27** ± 7 13 ± 1

24 ± 2 28 ± 3

M.F. Hossain et al. / Field Crops Research 131 (2012) 97–109

NO3-N leached (kg N ha-1 d-1)

0.8

105

Dhamrai 2008

Conv Eco

0.6

0.4

0.2

0 10-16Feb

17-24 Mar

25-2 Apr

NO3-N leached (kg N ha-1 d-1)

0.8

3-21 Mar

22-5 Apr

6-19 Apr

20-4 May

5-4 Jun

Conv

Daulatpur 2008

Eco 0.6

0.4

0.2

0 8-15 Feb

16-22 Feb

23-28 Feb

NO3-N leached (kg N ha-1 d-1)

0.8

1-20 Mar

21-4 Apr

4-19 Apr

20-4 May

5-6 Jun

Conv

Gabtali 2008

Eco 0.6

0.4

0.2

0 24-31Jan 1-6 Feb

7-13 Feb 14-20 Feb 21-28 Feb 1-7 Mar 8-23 Mar 24-4 Apr 5-19 Apr 20-5 May 6-21 May

NO3-N leached (kg N ha-1 d-1)

0.8

Conv

Shibganj 2008

Eco 0.6

0.4

0.2

0

24-31 Jan

1-6 Feb

7-13 Feb

14-20Feb 21-1 Mar 2-17 Mar

Duration

18-3 Apr

4-19 Apr

20-5 May

6-5 Jun

Fig. 3. Distribution of nitrate N leaching by resin bag method used during Boro rice growing season in the year 2008.

Gabtali in 2008 (P < 0.01) and at Shibganj in both 2007 and 2008 (P < 0.01, Table 7). In conventional farming, N removal varied from 18 to 27 kg ha−1 in 2007 and 17 to 27 kg ha−1 in 2008, whilst in ecological farming it was 13–22 kg ha−1 in 2007 and 22–28 kg ha−1 in 2008 (Table 7). Significant increases were only seen at Shibganj in 2008 (P < 0.05).

Significant differences in N removal were found between the two farming systems at Gabtali (P < 0.05) 2008 and at Shibganj in 2007 (P < 0.01). The weed N removal was higher in conventional farming systems, but not significantly, with some exceptions. Moreover, N loss due to weeds was greater in the conventional farming systems leading about 21–37% compared to ecological farming systems

106

M.F. Hossain et al. / Field Crops Research 131 (2012) 97–109

Table 8 Estimation of soil mineral N losses (leaching and gaseous) upon flooding (kg N ha−1 ) in conventional and ecological farming fields, 2007 and 2008 Boro rice growing seasons. Farming system

Dhamrai 2007

Daulatpur 2007

2008

Conventional 43 (3)a 45 Mean 32 SE ±2 ±1 ±4 Ecological 26 35 (3)a 43 Mean SE ±1 ±2 ±7 Comparing between conventional and ecological farming fields, probability level 0.24 0.77 0.05

2008

Gabtali 2007

2008

Shibganj 2007

2008

45 (5)a ±2

34 ±4

45 (3)a ±1

39 ±9

38 (3)a ±2

37 (5)a ±2

36 ±5

42 (3)a ±2

36 ±11

50 (3)a ±2

0.20

0.75

0.87

0.12

0.20

SE: standard error. a Leaching loss analysed by resin bag within 12 days after flooding.

where it was about 16–34% of the total crop yield. Existing weed control management practices urgently need to improve to make crop production economically sustainable and viable.

dry season, the soil moisture status was favourable to aerobic N transformations. The highest amount of NO3 − -N was leached within 21 days after transplanting at all four sites (Fig. 3). Presumably this reflects the loss of NO3 − -N accumulated during the dry conditions at the commencement of the wet season by leaching (Sierra, 1992). However, towards the end of growing season, when farmers were not irrigating, the aerobic conditions favoured the production of NO3 − -N, causing increased leaching in the following wet season. The amount and duration of the NO3 − -N varied at different sites, but was not significant for both farming practices. In four research sites for both farming practices cumulative NO3 − -N losses were measured during the Boro cropping season in 2008. These were from 22 and 23 kg N ha−1 at Dhamrai, 27 and 26 kg N ha−1 at Daulatpur, 19 and 20 kg N ha−1 at Gabtali, and 18 kg N ha−1 at Shibganj for both conventional and ecologically managed farming fields, respectively (Fig. 4). There were no significant differences for N leached from the conventional fields (18–22) compared to the ecological farming fields (18–26). Nitrogen losses due to leaching were measured relatively to fertiliser losses and losses upon flooding. However, NO3 − -N leaching from organic crop rotations are less due to the effects of location, manure and catch crop (Askegaard et al., 2005). These results suggest that the dominant loss pathways are gaseous.

3.2.3. Fertiliser N loss Low recovery efficiency of applied N potentially leads to high losses of fertiliser N (Table 6; Hossain et al., 2005). Large amounts of applied urea N losses indicated poor AE (N) achieved by conventional farming resulting from mismanagement of applied N. Nitrogen losses in conventional farming varied from 30 to 96 kg ha−1 in 2007 and from 38 to 133 kg ha−1 in 2008 and increased significantly in Dhamrai (P < 0.05) and Gabtali (P < 0.001) in 2007 (Table 8). 3.2.3.1. Loss of N upon flooding and leaching. In conventional farming fields, soil mineral N losses accounted for 32–45 kg ha−1 in 2007 and for 38–45 kg ha−1 in 2008, whilst in ecological fields they were 26–43 kg ha−1 in 2007 and 35–50 kg ha−1 in 2008 (Table 8). In 2008, soil mineral N content increased in all four research sites and for both farming systems, although the increases were only significant at Dhamrai for conventional farming (P < 0.01) and at Shibganj for ecological farming (P < 0.05). Significant differences between the two farming practices were only found at Dhamrai 2007 (P < 0.05; Table 8). Soil mineral N content was comparatively higher in the ecological farming than conventional farming systems. After the monsoon in 2007 and with the start of the Table 9 Nitrogen budgets (kg N ha−1 ) for Boro rice cropping seasons in 2007 and 2008. Dhamrai

Daulatpur

2007 CF Inputs Chemical fertiliser Organic manure Atmospheric N deposition Dry deposition Wet deposition Sedimentation Irrigation water (pump) Biological N fixation Total N inputs Outputs Grain Straw Weeds Loss upon irrigation Leaching Fertiliser N loss Total N outputs N balance

2008 EF

CF

Gabtali

2007 EF

CF

2008 EF

CF

Shibganj

2007 EF

CF

2008 EF

CF

2007 EF

CF

2008 EF

CF

EF

123 –

44

175 35

– 37

119 29

– 68

92 5

– 5

39 71

– 103

108 12

– 15

51 69

– 65

58 36

– 86

<1 4 10 4 10 151

<1 4 10 5 10 73

1 6 10 3 10 240

1 6 10 3 10 57

<1 4 10 4 10 176

<1 4 10 3 10 95

1 7 10 3 10 119

1 7 10 3 10 26

<1 4 – 8 10 132

<1 4 – 7 10 124

1 6 – 7 10 144

1 6 – 6 10 38

<1 3 – 8 10 131

<1 3 – 7 10 85

1 5 – 7 10 117

1 5 – 7 10 109

−69 −22 −24 −32a – −84 −231

−40 −12 −19 −26a – −97 −97

−68 −54 −25 −43 −22 −133 −345

−78 −32 −23 −35 −23 – −191

−54 −16 −26 −45a – 96 −237

−56 −10 −22 −43a

−59 −32 −22 −37 −26 – −176

−52 −51 −26 −21 −18 −22 −34a −36a – – −30 – −160 −130

−68 −38 −17 −45 −19 −67 −254

−69 −41 −26 −42 −20

−131

−75 −34 −27 −45 −27 −77 −285

−198

−54 −25 −27 −39a – −35 −180

−46 −20 −13 −36a – – −115

−57 −33 −24 −38 −18 −38 −208

−53 −36 −28 −50 −18 – −185

−80

−24

−105

−124

−61

−36

−157

−140

−110

−160

−39

−30

−91

−76

−28

−6

CF: conventional farming; EF: ecological farming. a Parenthesis indicates loss upon irrigation estimated including denitrification, volatilisation and leaching losses.

NO3-N leached (kg N ha-1)

M.F. Hossain et al. / Field Crops Research 131 (2012) 97–109

30 25 20 15 10 5 0

NO3-N leached (kg N ha-1)

0 30

20

80

100

Daulatpur Cumulative Leaching

Conv Eco

60

120

20 15 10 5 0 20

30 NO3-N leached (kg N ha-1)

40

25

0

40

60

80

Gabtali Cumulative Leaching

25

100

120

Conv Eco

20 15 10 5 0 0

20

30 NO3-N leached (kg N ha-1)

Conv Eco

Dhamrai Cumulative Leaching

40

60

80

100

Shibganj Cumulative Leaching

Conv Eco

25

120

20 15 10 5 0 0

20

40

60 80 Duration (days)

100

120

Fig. 4. Cumulative nitrate N leaching by resin bag method used during Boro rice growing season in the year 2008.

3.3. Nitrogen budgets Nitrogen budgets were constructed in Boro rice fields for both farming systems during two rice growing seasons for the four sites (Table 9). The budgets were negative for both farming systems in both years. Overall, ecological farming produced a less negative balance in both years (−6 to −36 kg N ha−1 in 2007 and −76 to −160 kg N ha−1 in 2008) than conventional farming (−28 to −80 kg N ha−1 in 2007 and −91 to −157 kg N ha−1 in 2008). The difference between conventional and ecological management outlines the high losses of fertiliser N under conventional management. 4. Discussion and conclusion On average, the inputs of N were lower under ecological than conventional farming. The budgets highlighted that even on

107

conventional fields farmers added up to two thirds of their fertiliser as organic manure. However, the amount of organic matter added to soil was not the same in all sites; it varied annually for both farming systems. The N budgets difference between the two farming systems reflects the high losses of fertiliser N under conventional management, indicating that ecological farming is more efficient. These fertiliser N losses stress the low agronomic efficiency of nitrogen fertiliser (Timsina et al., 2001; Hossain et al., 2005). Fertiliser broadcast onto the soil at transplanting is susceptible to loss by volatilisation or denitrification (Freney et al., 1990). The results of net immobilisation processes of uptake by aquatic weeds and green algae are higher in ecological than in conventional farming systems (Craswell and De Datta, 1980). They also cited that 10–25% of applied nitrogen could be immobilised by algae. Vlek et al. (1980) found that up to 40% of 15 N-labelled urea-nitrogen applied to rice in greenhouse pots could be immobilised by algae growing in the floodwater. Furthermore, studies at a nearby experimental site showed that broadcast soluble N fertilisers increased the growth of algae, particularly green algae, which do not fix nitrogen (Roger et al., 1980). Other immobilising agents might have been small aquatic weeds, which immobilised only 2–3% of broadcast urea 15 N in a similar experiment (Savant et al., 1982). The loss may therefore have been due to ammonia volatilisation or denitrification, particularly since the latter loss mechanisms involve nitrogen in mineral form in the soil; however lacking of soil extractions remind in the soil mineral nitrogen pool at harvest. Fertiliser losses were the predominant source of N loss under conventional management. These losses were in fact inferred from data on fertiliser N use efficiency. Given their significance these losses should be confirmed by measurement. It is important to recognise that large N losses seem to be a symptom of some other constraint – rather than poor N fertiliser management (Hossain et al., 2005). Furthermore, they indicate that yields are limited by a factor other than nitrogen. Applying N fertiliser does not solve P, K, Zn, S and other deficiencies (Eric Smaling, personal communication). Crop establishment, plant density, water and pest management are the key factors responsible for N losses. In addition the lack of agronomic performance may be due to poor seed and transplanted seedling quality, varieties that do not respond to N, or low radiation (Hossain et al., 2005). The key to improving agricultural production of rice appears to be isolating and overcoming those factors that limit grain formation. Until this can be done, there is a strong case for advising farmers at risk of losing yield to reduce fertiliser N inputs who are. However, negative N budgets can also impair the resource sustainability of agriculture through soil degradation and soil mining, resulting in a decline of soil fertility (FAO, 2003). These N budgets can be used to improve or design new technologies that tackle soil fertility management problems and also can help improve the financial performance of the farmers. Because from an economic stand point, soil fertility decline relates to short-term economic consideration of households, insecure climate and market environment, poor property rights, limited infrastructure and risk management as well. In order to tackle effectively the different problems in soil fertility decline, the integration of disciplines is a prerequisite, as the integration of formal science and farmers’ knowledge (FAO, 2003). The use of crop residues and weeds is an important part of nutrient management in rice based cropping systems in Bangladesh. Farmers did not incorporate straw because crop residues have a high value within the farming system as a source of fodder and fuel. Fuel use is linked to income and is fundamental to welfare. Fuel shortages can have serious implications for the nutrition of poorer farmers. These farmers use cowdung, rice and vegetable straw, and jute stick for fuel instead of using them as organic manure for input in to soil. They spend money on fuel instead of food. Larger

108

M.F. Hossain et al. / Field Crops Research 131 (2012) 97–109

farmers may allow those who work for them to keep both crop residues and animal manure as part of payment for services supplied. Therefore, N removal in straw was high. The harvested grain and straw jointly removed 70–122 kg N ha−1 from conventional farming and 52–110 kg N ha−1 from ecological farming, representing a total N removal of approximately 35–42% and 53–60%, respectively. During early crop growth, weeds were used as farm animal feed, or composted, or incorporated into the field. At Gabtali and Shibganj, this amount varied between 0 and 100%, but in general only up to 40% (approximately 7–15 kg N ha−1 ) of weeds were incorporated into the soil for the following T. Aman rice crop. Farmers at Dhamrai and Daulatpur, being closer to Dhaka, tended to sell surplus crop residues and weeds to residents in the capital where a farm animal’s fodder is in high demand. In addition to their importance in the N budget, weeds also pose major constraints to crop production and yields. Not only do weeds reduce the yields from rice plants by competing for space, nutrients and light, but weed seeds may also contaminate the eventual harvest. In irrigated sites, with good water control, keeping the fields flooded after planting tended to kill many types of weeds and also slowed the growth of others. However, the measured N budgets have also highlighted that under the common cropping system (Table 1) substantial amounts of mineral N accumulated during the transition period (Table 1) after harvest to Rabi crop and planting of Boro crop. This nitrate may be lost very quickly during land preparation and subsequent irrigation for boro (Tiedje, 1988). The soil mineral N losses varied between 32–50 kg ha−1 (13–21% of total N outputs) in conventional and 26–53 kg ha−1 (22–33% of total N outputs) in ecological farming fields. Furthermore, more nitrates accumulated under conventional management in 2007 and under ecological management in 2008. This increased level of nitrate accumulation may reflect on the increased N mineralisation potential in the same soils under ecological management (White et al., 1999). It may be possible to reduce these losses by altering fallow management or rice crop establishment. If land is prepared by puddling, differing types of fallow management have shown that the use of catch crop or weedy fallow can be used to retain mineral N, with the green manure (GM) being incorporated during land preparation (George et al., 1994). Alternatively, changing rice crop establishment from transplanting, which requires frequent irrigation and puddling, to dry surface seeding without puddling and infrequent irrigation may enable the rice plant to recover this nitrate directly. In addition, catch crops reduced NO3 − -N leaching by 30–38%, on sandy soils (Askegaard et al., 2005). Further to the benefit in conserving N fast growing tropical legumes can accumulate more than 80 kg N ha−1 in 45 days to meet the nutritional requirement of high yielding cultivars (Singh et al., 1991). Rice yield responses exceeding 2 t ha−1 are possible from 50 to 60 days old GM incorporation into soil (Singh et al., 1991). Sesbania cannabina as a GM was as effective as inorganic N sources on equal N basis for flooded rice (Singh et al., 1991). Obviously, many socio-economic and environmental factors determine whether an individual farmer will benefit from conserving this nitrogen and the most appropriate way to do so. These budgets however, indicate the magnitude and mechanisms of the various losses. An understanding of these losses and their consequences is important as a basis for considering management changes.

Acknowledgements The authors wish to acknowledge that this publication is an output from a project funded by the UK Department for

International Development (DFID) for the benefit of developing countries. The views expressed are not necessarily those of DFID. IACR-Rothamsted receives grant-aided support from UK Biotechnology and BBSRC. The authors also like to thanks to Professor Dr. Achim Dobermann, University of Nebraska, USA and Professor Dr. Ir. E.M.A. Smaling, ITC, The Netherlands for reviewing and editing on earlier version of the manuscript.

References Anderson, J.M., Ingram, J.S.I., 1993. Tropical Soil Biology and Fertility. A Handbook of Methods, second ed. CAB International, Wallingford, Oxon, UK. Askegaard, M., Olesen, J.E., Kristensen, K., 2005. Nitrate leaching from organic arable crop rotations: effects of location, manure and catch crop. Soil Use Manage. 21, 181–188. Atkins, D.H.F., Sandalls, J., Law, D.V., Hough, A.M., Stevenson, K., 1986. The measurement of nitrogen dioxide in the outdoor environment is using passive diffusion samplers. In: United Kingdom Atomic Energy Authority Report AERE R-12133. Harwell Laboratory, Oxfordshire. BARC (Bangladesh Agricultural Research Council), 2005. Fertiliser Recommendation Guide. BARC, Farmgate, Dhaka, Bangladesh. Brammer, H., 1983. Agriculture and food production in polders areas. A case study in Bangladesh. Water Int. 8, 74–81. Buresh, R.J., Woodhead, T., Shepherd, K.D., Floredies, E., Cabangon, R.C., 1989. Nitrate accumulation and loss in a mungbean lowland rice cropping systems. Soil Sci. Soc. Am. J. 53, 477–482. Campbell, W.G., 1988. Measurements of nitrogen dioxide concentrations at rural sites in the United Kingdom using diffusion tubes. Environ. Pollut. 55, 251–270. Cassman, K.G., De Datta, S.K., Amarante, S.T., Liboon, S.P., Samson, M.I., Dizon, M.A., 1996. Long-term comparison of the agronomic efficiency and residual benefits of organic and inorganic nitrogen sources for tropical lowland rice. Exp. Agric. 32, 427–444. Cassman, K.G., Krofff, M.J., Gaunt, J.L., Peng, S., 1993. Nitrogen use efficiency of rice reconsidered: what are the key constraint. Plant Soil. 155/156, 359–362. Craswell, E.T., De Datta, S.K., 1980. Recent developments in research on nitrogen fertilisers for rice. In: IRRI Res. Paper Series 49, p. 11. FAO, 2003. In: Roy, R.N., Misra, R.V., Lesschen, J.P., Smaling, E.M. (Eds.), Assessment of Soil Nutrient Balance: Approaches and Methodology. FAO Fertilizer and Plant Nutrition Bulletin no. 14, FAO of the United Nations, Rome, Italy. Freney, J.R., Trevitt, A.C.F., De Datta, S.K., Obcemea, W.N., Real, J.G., 1990. The independence of ammonia volatilisation and denitrification as nitrogen loss processes in flooded rice fields in the Philippines. Biol. Fert. Soils 9, 31–36. Gami, S.K., Ladha, J.K., Pathak, H., Shah, M.P., Pasuquin, E., Pandey, S.P., Hobbs, P.R., Joshy, D., Mishra, R., 2001. Long-term changes in yield and soil fertility in a twenty-year rice-wheat experiments in Nepal. Biol. Fert. Soils 34, 73–78. George, T., Ladha, J.K., Garrity, D.P., Torres, R.O., 1994. Legumes as nitrate catch crop during the dry to wet transition in low land rice cropping systems. Agron. J. 86, 267–273. Ghaley, B.B., Høgh-Jensen, H., Christiansen, J.L., 2010. Recovery of nitrogen fertilizer by traditional and improved rice cultivars in the Bhutan Highlands. Plant Soil 332, 233–246. Goulding, K.W.T., Bailey, N.J., Bradbury, N.J., Hargreaves, P., Howe, M., Murphy, D.V., Poulton, P.R., Willison, T.W., 1998. Nitrogen deposition and its contribution to nitrogen cycling and associated soil processes. New Phytol. 139, 39–58. Greenland, D.J., 1997. The Sustainability of Rice Farming. Cab International/IRRI, Wallingford, Oxon, UK/Manila, The Philippines, p. 79. Hall, D.J., 1986. The Precipitation Collector for use in the Secondary National Acid Deposition Network. Warren Spring Laboratory, Stevenage. Hargreaves, P.R., Leidi, A., Grubb, H.J., Howe, M.T., Mugglestone, M.N., 2000. Local and seasonal variations in atmospheric nitrogen dioxide levels at Rothamsted, UK, and relationships with meteorological conditions. Atmos. Environ. 34, 843–853. Hossain, M.F., 2001. The nitrogen economy of rice based cropping systems in Bangladesh. Ph.D. Thesis. Imperial College of Science, Technology and Medicine, University of London, UK. Hossain, M.F., Elahi, S.F., Khondaker, M., 2003. A comparison between ecological and conventional rice farming systems in Bangladesh. Diffusion Pollution Conference Dublin, Ireland. . 8C Ecology, 57–64. Hossain, M.F., White, S.K., Elahi, S.F., Sultana, N., Choudhury, M.H.K., Alam, Q.K., Rother, J.A., Gaunt, J.L., 2005. The efficiency of nitrogen fertiliser for rice in Bangladeshi farmers’ fields. Field Crops Res. 93, 94–107. Ivens, W.P.M.F., Draaijers, G.P.J., Bleuten, W., 1988. Atmospheric nitrogen deposition in a forest next to an intensively used agricultural site. In: Mathy, P., Reidel, D., Dordrecht, J.P. (Eds.), Air Pollution and Ecosystems. , pp. 536–541. Kundu, D.K., Ladha, J.K., 1999. Sustaining productivity of lowland rice soils: issues and options related to N availability. Nutr. Cycl. Agroecosyst. 53, 19–33. 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. Biol. Fert. Soils 42, 72–82. Ladha, J.K., Dawe, D., Pathak, H., Padre, A.T., Yadav, R.L., Singh, B., Singh Yadvinder, Singh, Y., Singh, P., Kundu, A.L., Sakal, R., Ram, N., Regmi, A.P., Gami, S.K., Bhandari, A.L., Amin, R., Yadav, C.R., Bhattarai, E.M., Das, S., Aggarwal, H.P., Gupta,

M.F. Hossain et al. / Field Crops Research 131 (2012) 97–109 R.K., Hobbs, P.R., 2003. How extensive are yield declines in long term rice-wheat experiments in Asia. Field Crops Res. 81, 159–180. Lane, P.W., Payne, R.D., 1998. Genstat for Windows: An Introductory Course. Version Genstat 5 Release 4.1, 4th ed. Statistics Department, IACR-Rothamsted, Harpenden, Herts, UK. Langeveld, J.W.A., Overbosch, G.B., 1996. Estimating nutrient surplus and nutrient use efficiency from farm characteristics. An application to private farms in two districts in Poland. Fert. Res. 45, 221–233. Ma, Q., Yu, W.-T., Shen, S.-M., Zhou, H., Jiang, Z.-S., Xu, Y.-G., 2010. Effects of fertilization on nutrient budget and nitrogen use efficiency of farmland soil under different precipitations in Northeastern China. Nutr. Cycl. Agroecosyst. 88, 315–327. Makarim, A.K., Sismiyati, R., Ponimin, P.W., 1994. Nitrogen requirement of irrigated rice at different growth stages. In: Berge, H.F.M., Woperies, M.C.S., Shin, J.C. (Eds.), Research Proceedings: Nitrogen Economy of Irrigated Rice: Field and Simulation Studies. SAR, pp. 70–82. Mian, A.M.J., Blume, H.P., Bhuiyan, Z.H., Eaqub, M., 1991. Water and nutrient dynamics of a paddy soil of Bangladesh. Z. Pfl. -Em. Bodenk 154, 93–99. Palmes, E.D., Gunninson, A.F., DiMattio, J., Tomczyk, C., 1976. Personal sampler for nitrogen dioxide. Am. Ind. Hyg. Assoc. J. 37. Perkauskas, D., Mikelinshiene, A., 1998. Evaluation of SO2 and NO2 concentration levels in Vilnius (Lithuania) using passive diffusion samplers. Environ. Pollut. 102 (S1), 249–252. Price, L.M.L., Balasubramanian, V., 1998. Securing the future of intensive rice systems: a knowledge-intensive resource management and technology approach. In: Dowling, N.G., Greenfield, S.M., Fischer, K.S. (Eds.), Sustainability of Rice in the Global Food System. , pp. 193–203. Reddy, K.R., Patrick, W.H., 1979. Nitrogen fixation in flooded rice soil. Soil Sci. 128, 80–86. Regmi, A.P., Ladha, J.K., Pathak, H., Pasquin, E., Bueno, C., Dawe, D., Hobbs, P.R., Joshy, D., Maskey, S.L., Pandey, S.P., 2002. Yield and soil fertility trends in a 20-year rice-rice-wheat experiment in Nepal. Soil Sci. Soc. Am. J. 66, 857–867. Roger, P.A., Kulasooriya, S.A., 1980. Blue Green Algae and Rice. International Rice Research Institute, Los Banos, Philippines, p. 112. Roger, P.A., Kulasooriya, S.A., Tirol, A., Craswell, E.T., 1980. Deep placement: a method of nitrogen fertiliser application compatible with algal nitrogen fixation in wetland rice soils. Plant Soil 57, 137–142. Rother, J.A., Whitton, B.A., 1989. Nitrogen activity of blue-green algae on seasonally flooded soils in Bangladesh. Plant Soil 113, 47–52. Savant, N.K., De Datta, S.K., Craswell, E.T., 1982. Distribution patterns of ammonium nitrogen and 15N uptake by rice after deep placement of urea supergranules in wetland soils. Soil Sci. Soc. Am. J. 46, 567–573. Sierra, J., 1992. Relationship between mineral N content and N mineralisation rate in disturbed soil samples incubated under field and laboratory conditions. Aust. J. Soil Res. 30, 477–492. Singh, Y., Khind, C.S., Singh, B., 1991. Efficient management of leguminous green manures in wet land rice. Adv. Agron. 45, 175–189. Slak, M.F., Commaganac, L., Lucas, S., 1998. Feasibility of national nitrogen balances. Environ. Pollut. 102, 235–240. Smaling, E.M.A., 1998. Nutrient balances as indicators of productivity and sustainability in sub-Saharan African agriculture. Agric. Ecosyst. Environ. 71, 235 (Special issue).

109

Smaling, E.M.A., Fresco, L.O., 1993. A decision-support model for monitoring nutrient balances under agricultural land use (NUTMON). Geoderma 60, 235–256. Smaling, E.M.A., Oenema, O., 1997. Estimating nutrient balances in agroeco-systems at different scales. In: Lal, R., Blume, W.H., Valentine, C., Stewart, B.A. (Eds.), Methods for Assessment of Soil Degradation. Advances in Soil Sciences. CRC Press, NY, pp. 229–252. Smaling, E.M.A., Stoorvogel, J.J., Windmeijer, P.N., 1993. Calculating soil nutrient balances in Africa at different scales. II District scales. Fert. Res. 35, 237–250. Song, C., Wang, E., Han, X., Stirzaker, R., 2011. Crop production, soil carbon and nutrient balances as affected by fertilisation in a Mollisol agroecosystem. Nutr. Cycl. Agroecosyst. 89, 363–374. Stoorvogel, J.J., Smaling, E.M.A., Janssen, B.H., 1993. Calculating soil nutrient balances in Africa at different scales. 1 Supra-national scale. Fert. Res. 35, 227–235. Tiedje, J.M., 1988. Ecology of denitrification and dissimilatory nitrate reduction to ammonium. In: Zehnder, A.J.B. (Ed.), Biology of Anaerobic Microorganisms. John, W., Sons, NY, pp. 79–244. 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 Res. 72, 143–161. Timsina, J., Panaullah, G.M., Saleque, M.A., Ishaque, M., Pathan, A.B.M.B.U., Quayyum, M.A., Connor, D.J., Shah, P.K., Humphreys, E., Meisner, C.A., 2006a. Nutrient uptake and apparent balances for rice-wheat sequences: I. Nitrogen. J. Plant Nutr. 29, 137–156. Timsina, J., Quayyum, M.A., Connor, D.J., Saleque, M., Haq, F., Panaullah, G.M., Jahan, M.A.H.S., Begum, R.A., 2006b. Effect of fertilizer and residue management on total productivity, soil fertility, and N-use efficiency of intensified rice–wheat systems. Int. J. Agric. Res. 1, 41–52. Triol Padre, A., Ladha, J.K., Singh, U.E., Laureles, E., Punzalan, G., Akita, S., 1996. Grain yield performance of rice genotypes at suboptimal levels of soil N as affected by N uptake and utilisation efficiency. Field Crops Res. 46, 127–143. Vitousek, P.M., 1994. Beyond global warming: ecology and global change. Ecology 75, 1861–1876. Vlek, P.L.G., Stumpe, J.M., Byrnes, B.H., 1980. Urease activity and inhibition in flooded soil systems. Fert. Res. 1, 191–202. White, S.K., Hossain, M.F., Sultana, N., Elahi, S.F., Chowdhury, M.H.K., Sarker, S., Alam, Q.K., Fortune, S., Rother, J., Gaunt, J.L., 1999. Low input ecological rice farming in Bangladesh. In: Proceedings for British Soil Science Society Meeting, September 16–17. Witt, C., Dobermann, A., Abdulrachman, S., Gines, H.C., Wang, G., Nagarajan, R., Satawatananont, S., Son, T.T., Tan, P.S., Le Van Tiem, Simbahan, G.C., Olk, D.C., 1999. Internal nutrient efficiencies of irrigated lowland rice in tropical and subtropical Asia. Field Crops Res. 63, 113–138. Yoneyama, T., Lee, K.K., Yoshida, T., 1977. Decomposition of rice residues in tropical soils IV. The effect of rice straw on fixation by heterotropic bacteria in some Phillippine soils. Soil. Sci. Plant Nutr. 23, 287–295. Zhao, L., Ma, Y., Liang, G., Li, S., Wu, L., 2009. Phosphorus efficacy in four Chinese long-term experiments with different soil properties and climate characteristics. Commun. Soil Sci. Plant Anal. 40, 3121–3138.