1 Introduction

The global population is expected to reach ~ 10.4 billion by 2050, which would be ~ 3 billion higher than in 2010 (Clay 2010; Tilman et al. 2011). It would require about 100–110% increase in food grain supply (Tilman et al. 2011) and ~ 60% increase in cereal production (FAO 2016). An estimated 80% of the increased food demand has to come from increase in crop yields (Alexandratos and Bruinsma 2012). Among different cereal-based cropping systems, rice–wheat practiced on ~ 13.5 million ha (Mha) in the Indo-Gangetic Plains (IGPs) of South Asia is the world’s largest production system (Timsina and Connor 2001). Rice (Oryza sativa L.) is a staple food for half of the world population (Godfray et al. 2010) and wheat (Triticum aestivum L.) is grown across a wide range of agro-climatic conditions in the world (Blumenthal et al. 2001). Rice ranks second, next to wheat in terms of area harvested globally, and ranks first in terms of calories per person (18% of total calories per person per day) (Mahajan and Gupta 2009). According to the Food and Agriculture Organization (FAO), humans consume about 80 kg rice and 66 kg wheat per capita per annum (FAO-STAT 2013).

In the Indian IGPs, rice–wheat system occupies ~ 10 Mha and accounts for 40% of the total food grains production of 281 million tones (MT) in the country (GOI 2018). However, the sustainability of rice–wheat cropping system is under threat because of low nutrient use efficiency (Benbi et al. 2011), emission of greenhouse gases (GHGs) (Sahai et al. 2011; Singh and Benbi 2020), and depletion of fossil fuel energy resources (Pishgar-Komleh et al. 2012). The agricultural sustainability of a production system besides productivity depends on is environmental and economic impacts (Helander and Delin 2004) of which improved nutrient use efficiency (Cassman et al. 2003) and energy efficacy (Khan et al. 2009) are significant (Dyer and Desjardins 2003; Dalgaard et al. 2001). Though crop productivity continues to increase with increasing rate of nutrient application and approaches optimum, nutrient use efficiency generally declines (Barbieri et al. 2008). The low nutrient use efficiency results in increased cost of crop production and environmental pollution (Fageria and Baligar 2005). Nitrogen use efficiency in cereals has been reported to be lowest in East Asia (32 kg kg−1) and highest in Africa (123 kg kg−1) (Dobermann and Cassman 2005). The imbalanced use of fertilizer N–P–K, e.g., 100–37–12 in India compared to 100–34–45 in the USA is responsible for lower nutrient use efficiency by crops in India (Norse 2003). Due to decreased nutrient use efficiency in rice–wheat cropping system, the use of chemical fertilizers for sustaining crop yields has been increasing that resulted in higher direct emission of GHGs from soils (Wassmann et al. 2009; Mandal et al. 2015) and indirect emissions associated with fertilizer manufacture and use (Schlesinger 1999; Singh et al. 2019a, 2019b). In a rice–wheat cropping system, chemical fertilizers are the major source of energy input accounting for 25–43% and 45–47% of total energy input in rice and wheat, respectively (Khan et al. 2009; Benbi 2018) with significant carbon (C) footprints (Shang et al. 2011; Ma et al. 2013; Singh and Benbi 2020). Therefore, for a climate-smart low C agriculture, rationalizing fertilizer application rates is of utmost importance. Balanced use of chemical fertilizers conjointly with organic manure is advocated to enhance C sequestration in soil (Benbi et al. 2016; Singh and Benbi 2018), but such management practices may also stimulate GHG emissions suggesting a trade-off between productivity and GHG emissions (Shang et al. 2011; Ma et al. 2013). In order to determine the extent of GHG emission per unit grain yield, the concept of greenhouse gas intensity (GHGI) was proposed (Mosier et al. 2006).

The assessment of indirect carbon dioxide equivalent (CO2e) emissions and energy flow in crop production requires quantitative appraisal of GHG emissions (Gu et al. 2012; Xia and Yan 2012). One such approach is a life cycle assessment (LCA) based on appraisal of the impact starting with the acquisition of raw material, processing during manufacture, and the use and the ultimate disposal of final product (Payraudeau and van der Werf 2005). LCA has been used to evaluate indirect GHG emissions related to application of chemical fertilizers in crop production (Duan et al. 2011; Gu et al. 2012; Xia and Yan 2012). It is therefore important to develop robust nutrient management practices to enhance productivity and nutrient use efficiency of a cropping system. Information on impact of long-term application of balanced fertilizers either alone or along with organic manure on nutrient use efficiency, energy efficacy, and net global warming potential (GWP) of a rice–wheat cropping system is scarce. Integrated nutrient management can bring efficiency in energy and nutrient use in crop production (Rautaray et al. 2012; Pishgar-Komleh et al. 2012; Meena et al. 2017). But, the increased energy input to maximize yield benefits may not always maximize monitory returns, because of increased crop production costs (Esengun et al. 2007; Saini et al. 2019). In contrast, increased production efficiency could be achieved with efficient energy use and may therefore, contribute towards economy, profitability, and sustainability of a cropping system (Ganajaxi et al. 2011; Tzanakakis et al. 2012).

The improvement in energy use efficiency in agriculture could reduce GHG emission while making crop production eco-friendly and sustainably cleaner enterprise (Pishgar-Komleh et al. 2012; Mi et al. 2016). Energy input–output relationships addresses energy depletion and climate change pressure related to crop production. Energy analysis unlike economic analysis is not influenced by fluctuation in the price of agricultural commodities (Jones 1989; Khan et al. 2009) and may therefore provide more realistic estimates of ecosystems’ economics and sustainability. Information on effect of balanced and imbalanced application of fertilizer nutrients with an objective of increasing nutrient use efficiency, while offsetting GHG emission and associated environmental and economic costs in rice–wheat cropping system, is scarce. We hypothesized that differential nutrient management including manure application would lead to variable impact on crops’ net ecosystem economic budget (NEEB) due to variable energy input and GHG emissions. More specifically, we aimed at investigating the effect of balanced and imbalanced application of fertilizer nutrients either alone or conjointly with farmyard manure (FYM) on (i) nutrient use efficiency, (ii) yield-scaled GHG emission to quantify magnitude of CO2e emission per unit grain yield, and (iii) energy relationships to reduce environmental and economic costs while sustaining yields of a rice–wheat cropping system.

2 Material and Methods

2.1 Experimental Site

A field experiment on rice (Oryza sativa L.)–wheat (Triticum aestivum L.) cropping system was established in 2008 at Regional Station, Gurdaspur (Latitude 30° 02′ 54.728 N, longitude 75° 25′ 34.280 E) of Punjab Agricultural University, Ludhiana, India. The region is characterized as hot humid, semi-arid, and subtropical with monsoonal climate. The average annual rainfall is ~ 713 mm, of which ~ 80% is received during the monsoon season extending from July to September and the rest 20% during the winter season. The field soil was sandy loam (sand = 56.2%, silt = 24.8%, and clay = 19.0%; Typic Ustorthents) and had pH = 7.53 (1:2 soil/water) and electrical conductivity (E.C.) = 0.29 dS m−1. The total organic C concentration at the surface (0–7.5 cm) and subsurface (7.5–15 cm) soil depths was 4.4 and 4.2 g kg−1, respectively (Benbi et al. 2016).

2.2 Treatment Details

The field experiment comprised nine treatments viz. unfertilized control (UF), application of nitrogen (N), phosphorus (P), and potassium (K) in different combinations (viz. N, NP, NK, PK, and NPK) and addition of farmyard manure (FYM), N + FYM, and NPK + FYM in rice–wheat cropping system. The treatments were replicated three times (plot size = 50 m2) in a randomized complete block design (RBD). The farmyard manure (15 Mg ha−1; moisture 40%) was applied at the time of field preparation for rice establishment in respective treatment plots. The FYM contained 250 g C kg−1 FYM and thus supplied 2.25 Mg C ha−1 year−1. Wheat (var. DBW-17) was sown at 100 kg ha−1 in rows 22.5 cm apart with seed-cum-fertilizer drill in the second week of November and harvested manually in the third week of April each year. The recommended fertilizer dose of 120 kg N, 26 kg P, and 25 kg K ha−1 was applied to wheat in NPK treatment. Half of the fertilizer N and full dose of fertilizer P and fertilizer K were applied to wheat at sowing, while the remaining half of N was applied with first irrigation. Wheat was adequately irrigated with groundwater and an irrigation of ~ 7.5 cm was applied as and when required depending on the visual inspection of the field.

Rice (var. PR-118) was established by transplanting the seedlings in wet plowed (puddling) fields. The field was pre-irrigated in the first fortnight of June and tilled to 10–15 cm depth twice thereafter at field capacity moisture to kill the weeds and to level the field for better puddling. It was inundated with 5–6 cm standing water and plowed twice under wet conditions by a tine cultivator, followed by planking with a wooden plank. About 25–30 days old rice seedlings were transplanted in a puddled field in rows 20 cm apart with plant-to-plant distance of 15 cm ensuring 33 plants m−2 area. Rice was transplanted in the second fortnight of June and harvested manually in the last week of October. In rice, fertilizer N at 120 kg N ha−1 was applied in three equal splits: at last puddling and 3 and 6 weeks after transplanting. The N, P, and K were applied through urea (46%N), single superphosphate (16% P2O5), and muriate of potash (60% K2O), respectively. In rice, water was kept standing continuously for 2 weeks after transplanting. Thereafter, irrigation was applied 2 days after ponded water has infiltrated in the soil.

2.3 Soil Sampling and Analysis

Soil samples from different treatments were collected from the surface (0–7.5 cm) and subsurface (7.5–15 cm) depths with a core sampler (inner diameter 7 cm). Each sample, a composite of four points in a plot was air-dried and passed through a 2-mm sieve and analyzed for mineral N, available P, and available K. Mineral N (nitrate-N; NO3−-N + ammonical-N; NH4+-N) was determined by extracting the soil samples with 2 M KCl, followed by N estimation by steam distillation of the extract using MgO and Devarda’s alloy (Mulvaney 1996). Available P in soil samples was determined by extracting with 0.5 M sodium bicarbonate (NaHCO3; pH = 8.5; 1:20: soil/extractant) (Olsen et al. 1954) and measuring P concentration in the extract colorimetrically (Murphy and Riley 1962). Available K in soil samples was extracted with 1 N ammonium acetate (NH4OAc, pH = 7.0) (Mervin and Peech 1950) followed by flame photometric determination.

2.4 Collection of Plant Samples and Analyses

Grain and straw samples collected after harvest of rice and wheat were dried at 50 °C in an oven. Nitrogen concentration in grains and straw samples was determined by the micro-Kjeldahl method. Phosphorus concentration was determined by digesting the plant samples in diacid (HNO3/HClO4; 4:1), followed by estimation on a spectrophotometer after developing yellow color using the molybedo-phosphoric acid method (Olsen and Sommers 1982). Potassium concentration was determined via flame photometer from the same extract that was used for determination of P. The N, P, and K uptake by rice and wheat grains and straw was calculated by multiplying the total crop dry matter yield with respective nutrient concentration.

2.5 Nutrient Use Efficiencies

The agronomic efficiency of applied N (AEN) was calculated from the ratio of difference in grain yield (kg ha−1) of fertilizer N treated and UF control to the amount (kg ha−1) of N applied (AN) (Eq. 1) (Fageria and Baligar 2005). Similarly, agronomic efficiencies of applied P (AEP) and K (AEK) were calculated from the ratio of difference in grain yield (kg ha−1) of fertilizer P and K treated and UF control to the amount (kg ha−1) of P (AP) and K applied (AK) using Eqs. 2 and 3. For the estimation of AEN, AEP, and AEK, the contribution of N, P, and K from FYM was not considered.

$$ {\mathrm{AE}}_{\mathrm{N}}\left(\mathrm{kg}\ \mathrm{grain}\ {\left(\mathrm{kg}\ \mathrm{N}\right)}^{-1}\right)=\frac{\mathrm{Grain}\ {\mathrm{yield}}_{\left(\mathrm{N}\right)}-{\mathrm{Grain}\ \mathrm{yield}}_{\left(\mathrm{UF}\right)}}{A_{\left(\mathrm{N}\right)}} $$
(1)
$$ {\mathrm{AE}}_{\mathrm{P}}\left(\mathrm{kg}\ \mathrm{grain}\ {\left(\mathrm{kg}\ \mathrm{P}\right)}^{-1}\right)=\frac{\mathrm{Grain}\ {\mathrm{yield}}_{\left(\mathrm{P}\right)}-{\mathrm{Grain}\ \mathrm{yield}}_{\left(\mathrm{UF}\right)}}{A_{\left(\mathrm{P}\right)}} $$
(2)
$$ {\mathrm{AE}}_{\mathrm{K}}\left(\mathrm{kg}\ \mathrm{grain}\ {\left(\mathrm{kg}\ \mathrm{K}\right)}^{-1}\right)=\frac{\mathrm{Grain}\ {\mathrm{yield}}_{\left(\mathrm{K}\right)}-{\mathrm{Grain}\ \mathrm{yield}}_{\left(\mathrm{UF}\right)}}{A_{\left(\mathrm{K}\right)}} $$
(3)

The physiological efficiency of N (PEN) was calculated as the ratio of difference in grain yield (kg ha−1) of fertilizer N applied and the UF control to that of the difference in N uptake (kg ha−1) by grain + straw from fertilizer applied and the UF control (Eq. 4) (Baligar and Fageria 2015). The physiological efficiency of P (PEP) and K (PEK) were estimated based on ratios of difference in grain yields and uptake of P and K (Eqs. 5 and 6).

$$ {\mathrm{PE}}_{\mathrm{N}}\left(\mathrm{kg}\ \mathrm{k}{\mathrm{g}}^{-1}\right)=\frac{\mathrm{Grain}\ {\mathrm{yield}}_{\left(\mathrm{N}\right)}-{\mathrm{Grain}\ \mathrm{yield}}_{\left(\mathrm{UF}\right)}}{{\mathrm{Uptake}}_{\left(\mathrm{N}\right)}-{\mathrm{Uptake}}_{\left(\mathrm{UF}\right)}} $$
(4)
$$ {\mathrm{P}\mathrm{E}}_{\mathrm{P}}\left(\mathrm{kg}\ \mathrm{k}{\mathrm{g}}^{-1}\right)=\frac{\mathrm{Grain}\ {\mathrm{yield}}_{\left(\mathrm{P}\right)}-{\mathrm{Grain}\ \mathrm{yield}}_{\left(\mathrm{UF}\right)}}{{\mathrm{Uptake}}_{\left(\mathrm{P}\right)}-{\mathrm{Uptake}}_{\left(\mathrm{UF}\right)}} $$
(5)
$$ {\mathrm{PE}}_{\mathrm{K}}\left(\mathrm{kg}\ \mathrm{k}{\mathrm{g}}^{-1}\right)=\frac{\mathrm{Grain}\ {\mathrm{yield}}_{\left(\mathrm{K}\right)}-{\mathrm{Grain}\ \mathrm{yield}}_{\left(\mathrm{UF}\right)}}{{\mathrm{Uptake}}_{\left(\mathrm{K}\right)}-{\mathrm{Uptake}}_{\left(\mathrm{UF}\right)}} $$
(6)

The recovery efficiency of applied N (REN) was calculated from the ratio of difference in uptake (kg ha−1) of N by grain + straw from fertilizer applied and the UF control to that of the amount (kg ha−1) of N applied (AN) using Eq. 7 (Baligar and Fageria 2015). Similarly, the recovery efficiency of P (REP) and K (REK) were estimated as a ratio of difference in P and K uptake and the amount of fertilizer P (AP) and K (AK) applied (Eqs. 8 and 9). For the estimation of REN, REP, and REK, the contribution of N, P, and K by FYM was not considered.

$$ {\mathrm{RE}}_{\mathrm{N}}\left(\%\right)=\frac{{\mathrm{Uptake}}_{\left(\mathrm{N}\right)}-{\mathrm{Uptake}}_{\left(\mathrm{UF}\right)}}{A_{\left(\mathrm{N}\right)}}\times 100 $$
(7)
$$ {\mathrm{RE}}_{\mathrm{P}}\left(\%\right)=\frac{{\mathrm{Uptake}}_{\left(\mathrm{P}\right)}-{\mathrm{Uptake}}_{\left(\mathrm{UF}\right)}}{A_{\left(\mathrm{P}\right)}}\times 100 $$
(8)
$$ {\mathrm{RE}}_{\mathrm{K}}\left(\%\right)=\frac{{\mathrm{Uptake}}_{\left(\mathrm{K}\right)}-{\mathrm{Uptake}}_{\left(\mathrm{UF}\right)}}{A_{\left(\mathrm{K}\right)}}\times 100 $$
(9)

The reciprocal internal use efficiency of applied N (RIUEN), P (RIUEP), and K (RIUEK) was calculated as the ratio of uptake (kg ha−1) of N, P, and K by grain to the grain yield (Mg ha−1) (Eqs. 10–12) (Setiyono et al. 2010).

$$ {\mathrm{RIUE}}_{\mathrm{N}}\left(\mathrm{kg}\ \mathrm{M}{\mathrm{g}}^{-1}\mathrm{grain}\ \mathrm{yield}\right)=\frac{{\mathrm{Uptake}}_{\left(\mathrm{N}\right)}}{\mathrm{Grain}\ \mathrm{yield}} $$
(10)
$$ {\mathrm{RIUE}}_{\mathrm{P}}\left(\mathrm{kg}\ \mathrm{M}{\mathrm{g}}^{-1}\mathrm{grain}\ \mathrm{yield}\right)=\frac{{\mathrm{Uptake}}_{\left(\mathrm{P}\right)}}{\mathrm{Grain}\ \mathrm{yield}} $$
(11)
$$ {\mathrm{RIUE}}_{\mathrm{K}}\left(\mathrm{kg}\ \mathrm{M}{\mathrm{g}}^{-1}\mathrm{grain}\ \mathrm{yield}\right)=\frac{{\mathrm{Uptake}}_{\left(\mathrm{K}\right)}}{\mathrm{Grain}\ \mathrm{yield}} $$
(12)

2.6 Crop Production Efficiency

The production efficiency of different treatments was calculated as a ratio of crop (rice, wheat) grain yield (kg ha−1) and crop duration (days) (Saini et al. 2019). The crop duration of 133 and 128 days, respectively was considered for estimating the production efficiency of rice and wheat (Eq. 13).

$$ \mathrm{Production}\ \mathrm{efficiency}\ \left(\mathrm{kg}\ \mathrm{h}{\mathrm{a}}^{-1}{\mathrm{day}}^{-1}\right)=\frac{\mathrm{Grain}\ \mathrm{yield}}{\mathrm{Crop}\ \mathrm{duration}} $$
(13)

2.7 Economics and Estimation of Economic Indices

The cost of cash inputs (CCI) was estimated as sum of expenditure incurred for the purchase of inputs (US$ ha−1), e.g., seed, fertilizers, FYM, insecticides, fungicides, herbicides, and labor cost for different field operations (Eq. 14) (Yang et al. 2020).

$$ \mathrm{CCI}\ \left(\mathrm{US}\$\ \mathrm{h}{\mathrm{a}}^{-1}\right)={\Sigma C}_{1,2,3,\dots \dots n} $$
(14)

where C1,2,3…. represents cost for different inputs and labor.

The gross returns (GR) were estimated as a product of crop (rice, wheat) grain yield (Mg ha−1) under different treatments and minimum support price (MSP) of produce (rice and wheat grains) fixed by Government of India (US$ Mg−1) (Eq. 15) (Yang et al. 2020).

$$ \mathrm{GR}\left(\mathrm{US}\$\ \mathrm{h}{\mathrm{a}}^{-1}\right)=\mathrm{Grain}\ \mathrm{yield}\times \mathrm{MSP} $$
(15)

The net returns (NR) were estimated as the difference in GR and CCI (Eq. 16) (Yang et al. 2020).

$$ \mathrm{NR}\left(\mathrm{US}\$\ \mathrm{h}{\mathrm{a}}^{-1}\right)=\mathrm{GR}-\mathrm{CCI} $$
(16)

The CCI, GR, and NR were computed by converting Indian rupee (INR) to US Dollar (US$), considering 1 US$ = 60 INR.

The gross returns above the fertilizer cost (GRAFC) were estimated as the difference in between GR and amount (A) of N, P, and K applied (kg ha−1) and price of N, P, and K (US$ kg−1) as per treatment (Eq. 17) (Yang et al. 2020).

$$ \mathrm{GRAFC}\left(\mathrm{US}\$\ \mathrm{h}{\mathrm{a}}^{-1}\right)=\left[\mathrm{GR}\right]-\left[{\mathrm{A}}_{\left(\mathrm{N}+\mathrm{P}+\mathrm{K}\right)}\times \mathrm{pric}{\mathrm{e}}_{\left(\mathrm{N}+\mathrm{P}+\mathrm{K}\right)}\right] $$
(17)

where A(N + P + K) represents amount of N + P + K applied and price(N + P + K) represents the market price of N + P + K. These estimations were based on market price of US$ 90, 210, and 233 Mg−1 for urea, diammonium phosphate, and muriate of potash, respectively.

Gross returns above the fertilizer cost for rice–wheat cropping system (GRAFCCS) were estimated as sum total of gross returns above fertilizer cost for rice (GRAFCR) and wheat (GRAFCW) (Eq. 18) (Singh and Benbi 2020).

$$ \mathrm{GRAF}{\mathrm{C}}_{\mathrm{C}\mathrm{S}}\left(\mathrm{US}\$\ \mathrm{h}{\mathrm{a}}^{-1}\right)=\mathrm{GRAF}{\mathrm{C}}_{\mathrm{R}}+\mathrm{GRAF}{\mathrm{C}}_{\mathrm{W}} $$
(18)

The economic efficiency was determined by dividing net returns (NR, US$ ha−1) from crop (rice, wheat) under different fertilizer treatments by crop duration (days) (Eq. 19) (Saini et al. 2019). The crop duration was kept same as for the estimation of production efficiency (Eq. 13).

$$ \mathrm{Economic}\ \mathrm{efficiency}\ \left(\mathrm{US}\$\ \mathrm{h}{\mathrm{a}}^{-1}{\mathrm{day}}^{-1}\right)=\frac{\mathrm{Net}\ \mathrm{returns}\ \left(\mathrm{NR}\right)}{\mathrm{Crop}\ \mathrm{duration}} $$
(19)

2.8 Energy Auditing and Estimation of Energy Indices

Energy input for different treatments was estimated from quantity of input (manure and fertilizers as per treatment) applied. The output energy was estimated from quantity of above-ground (grain plus straw) biomass produced. For estimating the energy value, quantities of input and output were converted to energy units by multiplying with their respective energy equivalents. In the agricultural production system, total input energy is classified as direct and indirect energy sources (Devasenapathy et al. 2009). Since the present study focused on balanced and imbalanced application of fertilizer nutrients under rice–wheat cropping system, the direct sources of energy viz. diesel, agricultural implements, human and animal power, etc., were not considered. Only the indirect energy sources that do not release energy directly, but dissipate energy in different processes viz. crop biomass (grain and straw), fertilizer N, P, and K and FYM, were considered to develop energy input–output relationships. The energy coefficients of 60.6 MJ (kg N)−1, 11.1 MJ (kg P2O5)−1, 6.7 MJ (kg K2O)−1, and 0.3 MJ (kg FYM)−1 were used for calculating energy input (expressed in MJ ha−1) in different treatments (Esengun et al. 2007). Energy input through the application of fertilizer N, P, and K, and FYM were summed (as per treatment) to estimate total input energy. The amount of energy produced from the grain and straw yield of rice and wheat was summed to estimate total output energy. The total energy output for each crop was obtained by multiplying grain and straw yields by their respective energy equivalents (14.7 MJ ha−1for rice grain, 15.7 MJ ha−1 for wheat grain, and 12.5 MJ ha−1for both rice and wheat straw) (Ozkan et al. 2004; Moradi and Azarpour 2011).

Different energy indices viz. energy productivity (EP), specific energy (SE), energy intensiveness (EI), energy ratio or energy use efficiency (ER), and the net energy gain (NEG) for rice and wheat were calculated using the following equations (Eqs. 20–26):

$$ \mathrm{Total}\ \mathrm{input}\ \mathrm{energy}\ \left(\mathrm{MJ}\ {\mathrm{ha}}^{-1}\right)=\sum \left({E}_{\mathrm{N}}+{E}_{{\mathrm{P}}_2{\mathrm{O}}_5}+{E}_{{\mathrm{K}}_2\mathrm{O}}+{E}_{\mathrm{FYM}}\right) $$
(20)
$$ \mathrm{Total}\ \mathrm{output}\ \mathrm{energy}\ \left(\ \mathrm{MJ}\ {\mathrm{ha}}^{-1}\right)=\sum \left({E}_{\mathrm{Grain}}+{E}_{\mathrm{Straw}}\right) $$
(21)
$$ \mathrm{Energy}\ \mathrm{productivity}\ \left({E}_{\mathrm{P}},\mathrm{kg}\ {\mathrm{MJ}}^{-1}\right)=\frac{Y_{\mathrm{Grain}+\mathrm{straw}}}{\mathrm{Total}\ \mathrm{input}\ \mathrm{energy}} $$
(22)
$$ \mathrm{Specific}\ \mathrm{energy}\ \left({S}_{\mathrm{E},}\mathrm{MJ}\ {\mathrm{kg}}^{-1}\right)=\frac{\mathrm{Total}\ \mathrm{input}\ \mathrm{energy}}{Y_{\mathrm{Grain}+\mathrm{straw}}} $$
(23)
$$ \mathrm{Energy}\ \mathrm{intensiveness}\ \left(\mathrm{EI},\mathrm{MJ}\ \mathrm{US}{\$}^{-1}\right)=\frac{\mathrm{Total}\ \mathrm{input}\ \mathrm{energy}}{\mathrm{CCI}} $$
(24)
$$ \mathrm{Energy}\ \mathrm{ratio}\ \left({E}_{\mathrm{R}}\right)=\frac{\mathrm{Total}\ \mathrm{output}\ \mathrm{energy}}{\mathrm{Total}\ \mathrm{input}\ \mathrm{energy}} $$
(25)
$$ \mathrm{Net}\ \mathrm{energy}\ \mathrm{gain}\ \left(\mathrm{NEG}\right)=\mathrm{Total}\ \mathrm{output}\ \mathrm{energy}-\mathrm{total}\ \mathrm{input}\ \mathrm{energy} $$
(26)

where EN, \( {E}_{{\mathrm{P}}_2{\mathrm{O}}_5} \), \( {E}_{{\mathrm{K}}_2\mathrm{O}} \), and EFYM represent energy input (MJ ha−1) through fertilizer N, P, K, and FYM, respectively, under rice–wheat cropping system. The EGrain and EStraw represent the energy output (MJ ha−1) through grain and straw, respectively, of rice and wheat. The total energy output through grain and straw biomass under rice–wheat cropping system is expressed as GJ ha−1 (1 GJ = 1000 MJ). The YGrain + straw is the grain + straw yield (each in kg ha−1) of rice and wheat.

2.9 Estimation of GHG Emission, Global Warming Potential, and GHG Intensity

The environmental impacts of different nutrient management practices under rice–wheat cropping system were assessed by calculating GWP using the LCA approach. The indirect GHG (CO2 and N2O) emissions due to fertilizer application under rice–wheat cropping system were estimated using CO2 equivalents (CO2e). Emission factors for GHG emission from different nutrient inputs such as 1.3 kg C (kg N)−1, 0.2 kg C (kg P2O5)−1, 0.15 kg C (kg K2O)−1, and 0.0075 kg C (kg FYM)−1 were used (Stout 1990; Lal 2004). The estimated CO2 and N2O emissions were converted to CO2e using GWP equivalent factors of 1, 265, and 25, respectively, for CO2, N2O, and CH4 for 100 years time frame (IPCC 2013) using Eq. 27. This means that emissions of 1 kg of CH4 and N2O, respectively, are equivalent to emissions of 25 and 265 kg of CO2 (IPCC 2007). The GHG emissions from fertilizer and manure application were calculated by multiplying the quantity of input with their respective emission coefficients (Lal 2004).

$$ \mathrm{GWP}\ \left(\mathrm{kg}\ {\mathrm{CO}}_{2\mathrm{e}}\ {\mathrm{ha}}^{-1}\right)=\left({\mathrm{CO}}_2\right)+\left({\mathrm{CH}}_4\times 25\right)+\left({\mathrm{N}}_2\mathrm{O}\times 265\right) $$
(27)

The CH4 emission during rice cultivation period was estimated using following equation based on the IPCC guidelines (tier 2) (IPCC 2006) (Eq. 28).

$$ {\mathrm{CH}}_4\ \mathrm{emission}\ \left(\mathrm{kg}\ {\mathrm{CH}}_4\ {\mathrm{ha}}^{-1}\right)=\mathrm{EF}\times t $$
(28)

where EF is the adjusted daily emission factor (kg CH4 ha−1 day−1) and t is the average rice cultivation period. Average rice cultivation period of 133 days was considered; the same duration was used for the estimation of production and economic efficiency.

The N2O emissions for N applied through chemical fertilizer and manure were estimated using following equation (Eq. 29).

$$ {\mathrm{N}}_2\mathrm{O}\ \mathrm{emissions}=\mathrm{N}\times {\mathrm{EF}}_1\times 44/28 $$
(29)

where N2O emissions (kg N2O ha−1) = N2O emissions from fertilizer N/FYM application under treated soils. N = N input through chemical fertilizer and FYM (kg N ha−1). Total N input through FYM was calculated by considering N concentration of 0.484% in FYM (Stout 1990). The EF1 = emission factor 0.01 for N2O emissions from N inputs, kg N2O-N kg−1 N input.

The GHGI was estimated by dividing GWP by crop (rice, wheat) grain yield (Eq. 30) and was expressed as kg CO2e kg−1 grain yield (Shang et al. 2011).

$$ \mathrm{GHGI}=\frac{\ \mathrm{GWP}}{\mathrm{Gain}\ \mathrm{yield}} $$
(30)

The carbon equivalent emissions (CEE) and carbon efficiency ratio (CER) of the fertilizer and manure application treatments were calculated using Eqs. 31 and 32 (Bhatia et al. 2005).

$$ \mathrm{CEE}\ \left(\mathrm{Mg}\ \mathrm{C}\ {\mathrm{ha}}^{-1}\right)=\mathrm{GWP}\times 12/44 $$
(31)
$$ \mathrm{CER}=\mathrm{grain}\ \mathrm{and}\ \mathrm{straw}\ \mathrm{yield}\ \mathrm{in}\ \mathrm{terms}\ \mathrm{of}\ \mathrm{C}\ \mathrm{of}\ \mathrm{the}\ \mathrm{rice}\ \mathrm{and}\ \mathrm{wheat}/\mathrm{CEE} $$
(32)

The C concentration in rice and wheat grains was assumed to be 400 g kg−1. Straw C yield of rice and wheat was estimated using C concentrations of 381 and 352 g kg−1, respectively (Majumder et al. 2007). The grain and straw C yields of rice and wheat were estimated using Eqs. 33 and 34.

$$ \mathrm{Grain}\ \mathrm{C}\ \mathrm{yield}\ \left(\mathrm{Mg}\ \mathrm{C}\ {\mathrm{ha}}^{-1}\right)=\mathrm{Crop}\ \left(\mathrm{rice}\ \mathrm{and}\ \mathrm{wheat}\right)\ \mathrm{grain}\ \mathrm{yield}\ \left(\mathrm{Mg}\ {\mathrm{ha}}^{-1}\right)\times \left(\frac{\mathrm{Grain}\ \mathrm{C}\ \mathrm{concentration}\ \left(\mathrm{g}\ {\mathrm{kg}}^{-1}\right)}{10}\right) $$
(33)
$$ \mathrm{Straw}\ \mathrm{C}\ \mathrm{yield}\ \left(\mathrm{Mg}\ \mathrm{C}\ {\mathrm{ha}}^{-1}\right)=\mathrm{Crop}\ \left(\mathrm{rice}\ \mathrm{and}\ \mathrm{wheat}\right)\ \mathrm{straw}\ \mathrm{yield}\ \left(\mathrm{Mg}\ {\mathrm{ha}}^{-1}\right)\times \left(\frac{\mathrm{Sraw}\ \mathrm{C}\ \mathrm{concentration}\ \left(\mathrm{g}\ {\mathrm{kg}}^{-1}\right)}{10}\right) $$
(34)

The total C yield of a rice–wheat cropping system was estimated as a sum of grain and straw yield of rice and wheat (Eqs. 35 and 36).

$$ \mathrm{Total}\ \mathrm{C}\ \mathrm{yield}\ \mathrm{of}\ \mathrm{rice}\ \left(\mathrm{Mg}\ \mathrm{C}\ {\mathrm{ha}}^{-1}\right)=\mathrm{Rice}\ \mathrm{grain}\ \mathrm{C}\ \mathrm{yield}\ \left(\mathrm{Mg}\ \mathrm{C}\ {\mathrm{ha}}^{-1}\right)+\mathrm{rice}\ \mathrm{straw}\ \mathrm{C}\ \mathrm{yield}\ \left(\mathrm{Mg}\ \mathrm{C}\ {\mathrm{ha}}^{-1}\right) $$
(35)
$$ \mathrm{Total}\ \mathrm{C}\ \mathrm{yield}\ \mathrm{of}\ \mathrm{wheat}\ \left(\mathrm{Mg}\ \mathrm{C}\ {\mathrm{ha}}^{-1}\right)=\mathrm{Wheat}\ \mathrm{grain}\ \mathrm{C}\ \mathrm{yield}\ \left(\mathrm{Mg}\ \mathrm{C}\ {\mathrm{ha}}^{-1}\right)+\mathrm{wheat}\ \mathrm{straw}\ \mathrm{C}\ \mathrm{yield}\ \left(\mathrm{Mg}\ \mathrm{C}\ {\mathrm{ha}}^{-1}\right) $$
(36)

2.10 Net Ecosystem Economic Budget

The economic feasibility of different fertilizer treatments in rice–wheat cropping systems was assessed through NEEB, which was calculated according to Zhang et al. (2015) using Eq. 37.

$$ \mathrm{NEEB}=\mathrm{GR}-\left(\mathrm{fertilizer}/\mathrm{manure}\ \mathrm{cost}\right)-\left(\mathrm{GWP}\ \mathrm{costs}\right) $$
(37)

where GR is the gross returns (US$ ha−1) calculated from the product of crop (rice + wheat) grain yield and MSP of rice and wheat (Eq. 7). The GWP costs were calculated based on the trade price of C (16.15 US$ Mg−1 CO2e) and GWP (Li et al. 2015; Zhang et al. 2015).

2.11 Statistical Analysis

Data on different parameters (grain yield, straw yield, nutrient use efficiency, nutrient concentration, energy and economic indices, etc.) were subjected to one-way analysis of variance (ANOVA) as per randomized complete block design (RBD) (Cochran and Cox 1966) using SPSS for Windows 16.0 (SPSS Inc., Chicago, USA). Mean separation for nutrient management treatments was performed using Duncan’s multiple range test (DMRT). The treatment means differences at p < 0.05 were considered statistically significant. The statistical significance of coefficient of determination (R2) was tested using F test. The R2 significant at p < 0.05 and p < 0.01 was marked as * and **, respectively.

3 Results

3.1 Nutrient Availability in Soil

The NO3−-N and NH4+-N concentrations in soil increased significantly with fertilizers and manure application (Table 1). Fertilizer N application either alone or together with P or K did not influence NO3−-N and NH4+-N concentration at both the surface and the subsurface soil depths. Application of FYM alone significantly increased NO3−-N and NH4+-N concentration compared with UF control. Conjoint application of NPK and FYM significant increased NO3−-N and NH4+-N concentration compared with their individual application. The available P did not differ significantly between NP and PK treatments. Similarly, available P concentration did not change significantly with NK compared with N alone. However, the application of NPK alone significantly increased P concentration compared with FYM alone. The conjoint application of NPK and FYM significantly increased available P concentration by ~ 15.2 and 24.7%, respectively, compared with alone NPK and FYM application in the surface soil layer. The corresponding increase in the sub-surface soil layer was ~ 17.9 and 28.9%, respectively. As compared with UF control, available P concentration increased by ~ 54.8 and 58.2%, respectively with NPK + FYM application in surface and subsurface soil layer. Available K in soil did not increase significantly with FYM alone, compared with UF control. Conversely, NPK alone application resulted in a significant increase in available K concentration, compared with UF control (Table 1).

Table 1 Effect of nutrient management practices on mineral N, available P and K in surface and sub-surface soil layers after 5 cycles of rice-wheat cropping in northwestern India
Full size table

3.2 Crop Yield and Production Efficiency

Rice grain yield increased significantly with the application of both organic and inorganic fertilizers, compared with UF control (Fig. 1). Significantly higher rice and wheat grain yield was recorded in plots receiving NPK and NPK + FYM, compared with imbalanced application of fertilizer nutrients. In the absence of fertilizer N, adding P and K alone significantly decreased the crop biomass (grain and straw), implying that N was a yield limiting factor. Application of NPK increased the yield responses with yield increase of 12% for rice and 3.2% for wheat as compared to N application only. Similarly, compared with PK alone, the NPK treatment increased rice grain yield by 48.2% and wheat grain yield by 111.2%. Fertilizer P application tended to increase rice and wheat grain yield by 6.5 and 2.2%, respectively, compared with NK alone. Fertilizer K application tended to significantly increase the grain yield of rice and wheat (comparing NPK with NP), indicating that K was also the yield limiting factor in rice–wheat cropping system. The rice grain yield with NPK + FYM application increased by 3 and 52%, respectively, compared with NPK and FYM alone. The NPK + FYM application resulted in 7 and 90% increase in wheat grain yield, compared with NPK and FYM alone, respectively. The combined application of NPK + FYM resulted in a significant increase in rice and wheat grain yield by 58 and 141%, respectively, than the UF control.

Fig. 1
figure 1

Effect of nutrient management practices on grain and straw yield of rice and wheat grown in sequence in northwestern India. Line bars indicate standard error from mean. Mean values for a crop followed by different letters are significantly different (p < 0.05) by Duncan’s multiple range test (DMRT)

Full size image

The production efficiency was significantly lower in UF control and increased with the application of chemical fertilizers (Fig. 2). Application of FYM alone did not significantly increase the production efficiency of rice. On the contrary, the residual effect of FYM application in wheat was significant. Unlike in wheat, the production efficiency did not differ significantly among NPK, N + FYM, and NPK + FYM treatments in rice. Conjoint application of NPK + FYM significantly increased the production efficiency of wheat by 3.0 kg ha−1 day−1 compared with the application of NPK alone.

Fig. 2
figure 2

Effect of nutrient management practices on production efficiency of rice and wheat in northwestern India. Line bars indicate standard error from mean. Mean values for a crop followed by different letters are significantly different (p < 0.05) by Duncan’s multiple range test (DMRT)

Full size image

3.3 Nutrient Uptake and Nutrient Use Efficiency

Nutrient (N, P, and K) uptake increased significantly with the application of manure and inorganic fertilizers, compared with UF control (Table 2). Fertilizer N application significantly increased N uptake by grain + straw of rice and wheat, compared with PK and UF control, although the two treatments did not differ significantly. Application of NPK and FYM alone significantly increased N uptake by rice and wheat, compared with UF control. Magnitude of increased N uptake with NPK application was higher for rice (by ~ 56.7%) and wheat (by ~ 85.3%), respectively, as compared to FYM alone. As compared to UF control, P uptake by rice increased significantly with NPK (by ~ 64.5%) and NPK + FYM application (by ~ 69.8%), although the two treatments did not differ significantly. The combined application of NPK + FYM increased the P uptake by rice by ~ 3.2 and 61.3%, respectively, compared with alone NPK and FYM application. In wheat, P uptake increased by ~ 7.8 and 96.4%, respectively, with NPK + FYM than the alone NPK and FYM application, respectively.

Table 2 Effect of nutrient management practices on N, P, and K uptake and increase over unfertilized (UF) control in rice and wheat grown in a sequence in northwestern India
Full size table

The K uptake by rice and wheat was significantly lower in PK, compared with NK treatment. However, balanced application (NPK) resulted in a significantly higher K uptake, compared with imbalanced fertilizer application. Although, FYM alone did not significantly influence K uptake, but when applied conjointly with fertilizers (N + FYM and NPK + FYM), K uptake by rice and wheat increased significantly. Over the UF control, percent increase in nutrient (N, P, and K) uptake was the lowest in PK and the highest in NPK + FYM (Table 2). Application of FYM alone also had a small effect on nutrient uptake by rice and wheat, compared with the application of inorganic fertilizers. When FYM was conjointly applied with fertilizers, percent increase in nutrient uptake increased substantially.

The AEN in rice and wheat responded significantly to fertilizer application (Table 3). Highest AEN of 21.4 kg grains (kg N)−1 in rice and 24.4 kg grains (kg N)−1 in wheat occurred in NPK treatment. A non-significant change in AEN in rice and wheat occurred with NP application, compared with N alone. However, it increased significantly by 6.1 kg grains (kg N)−1 in rice and 1.4 kg grains (kg N)−1 in wheat with NPK application, compared with UF control. The AEP and AEK were significantly increased with conjoint application of NPK + FYM. The REN in rice and wheat increased significantly by ~ 10.6 and 14.5%, respectively, with NPK + FYM application, compared with NPK alone. The REP and REK were significantly increased with conjoint application of NPK + FYM, compared with other treatments. The PEN increased significantly by 14.9 kg (kg N)−1 in rice and by 8.1 kg (kg N)−1 in wheat with NPK application, compared with N alone. The conjoint application of NPK + FYM did not significantly influence the PEN compared with NPK alone. In general, AEN and PEN were significantly higher for wheat than rice. The increase in AEN exhibited a significant linear increase in the production efficiency of a rice–wheat cropping system (Fig. 3), and the relationship was best described by a linear function (Eqs. 38 and 39).

Table 3 Effect of nutrient management practices on agronomic efficiency (AE), recovery efficiency (RE), and physiological efficiency (PE) of N, P, and K in rice and wheat grown in a sequence in northwestern India
Full size table
Fig. 3
figure 3

Relationship between crop production efficiency and agronomic efficiency of N in rice and wheat in northwestern India. Line bars indicate standard error from mean

Full size image
$$ \mathrm{Production}\ \mathrm{efficiency}\ \mathrm{of}\ \mathrm{rice}\ \left(\mathrm{kg}\ {\mathrm{ha}}^{-1}\ {\mathrm{day}}^{-1}\right)=0.98\ {\mathrm{AE}}_{\mathrm{N}}\ \left(\mathrm{kg}\ \mathrm{grain}\ \left(\mathrm{kg}\ \mathrm{N}\Big){}^{-1}\right)\ \right(\mathrm{kg}\ \mathrm{N}\ \mathrm{kg}+34.4,{R}^2=0.927\ast, p<0.01 $$
(38)
$$ \mathrm{Production}\ \mathrm{efficiency}\ \mathrm{of}\ \mathrm{wheat}\ \left(\mathrm{kg}\ {\mathrm{ha}}^{-1}\ {\mathrm{day}}^{-1}\right)=1.04\ {\mathrm{AE}}_{\mathrm{N}}\ \left(\mathrm{kg}\ \mathrm{grain}\ \left(\mathrm{kg}\ \mathrm{N}\Big){}^{-1}\right)\ \right(\mathrm{kg}\ \mathrm{N}\ \mathrm{kg}+19.2,{R}^2=0.938\ast, p<0.01 $$
(39)

The RIUEN varied between 10.2 and 12.4 kg Mg−1 rice grain and between 7.5 and 8.4 kg Mg−1 wheat grain (Fig. 4). Average across the treatments, RIUEN in rice (11.6 kg Mg−1) was higher by ~ 46% than wheat. Similarly, RIUEP and RIUEK in rice were higher by ~ 50.7 and 56.4%, respectively, than in wheat. Regardless of treatments, reciprocal internal use efficiencies followed an order RIUEN > RIUEK > RIUEP for both rice and wheat crops. In rice, RIUEN, RIUEP, and RIUEK were significantly lower in PK treatment and increased significantly with N application (N, NP, NK, and NPK). In rice, average PE were 78.3 kg grains per kg plant N, 218.8 kg grains per kg plant P, and 127.1 kg grains per kg plant K, equivalent to RIUE of 11.6 kg for N, 5.7 kg for P, and 10.4 kg for K to produce 1 Mg grain giving N/P/K ratio of 2.04:1:1.82. However, in wheat average PE were 121.4 kg grains per kg plant N, 254.3 kg grains per kg plant P, and 109.0 kg grains per kg plant K, equivalent to RIUE of 7.9 kg for N, 3.8 kg for P, and 6.7 kg for K to produce 1 Mg grain gave N/P/K ratio of 2.07:1:1.76 (Table 3, Fig. 4).

Fig. 4
figure 4

Effect of nutrient management practices on reciprocal internal use efficiency of nitrogen (RIUEN), phosphorus (RIUEP), and potassium (RIEUK) of rice and wheat in northwestern India. Line bars indicate standard error from mean. Mean values for a crop, followed by different letters are significantly different (p < 0.05) by Duncan’s multiple range test (DMRT)

Full size image

3.4 Greenhouse Gas Emission and Carbon Budget of Above-Ground Biomass

The indirect emission of CO2 and N2O due to fertilizer application in rice–wheat cropping system was the highest in NPK + FYM and the lowest in PK (Table 4). The GWP of a rice–wheat cropping varied between 10.4 and 13.6 Mg CO2e ha−1 year−1, the lowest in UF control, and the highest in NPK + FYM treatment. The GWP increased with NPK + FYM application by ~ 6.3 and 22.5%, respectively, compared with NPK and FYM. As compared with UF control, the GWP increased by ~ 27.9% with NPK + FYM application in rice–wheat/rice–wheat cropping system. The GHGI was significantly higher (by ~ 77.8%) in UF control, compared with NPK + FYM. The conjoint application of NPK + FYM significantly decreased the GHGI by 2.06 kg CO2e kg−1 grain yield (~ 35.8%), compared with FYM alone. A significant decrease in GHGI with NPK (by ~ 35.5%), compared with the application of FYM alone was observed. The CEE was significantly higher in manure applied plots (N + FYM, NPK + FYM), compared with other treatments. The effect of PK treatment on CEE of rice–wheat cropping system was at par with UF control.

Table 4 Effect of nutrient management practices on greenhouse gas (GHG) emission, global warming potential (GWP), greenhouse gas intensity GHGI), and carbon equivalent emission (CEE) under rice–wheat cropping system in northwestern India
Full size table

Fertilizer application significantly increased grain and straw C yield of both rice and wheat. Rice grain C yield increased significantly by 1.08 Mg C ha−1 (60%) with NPK application, compared with UF control (Table 5). NPK application increased wheat grain C yield by 2.25 Mg C ha−1 (67.6%) than the UF control. Application of NPK + FYM significantly increased the grain and straw C yield of rice and wheat, compared with the application of FYM alone. Total C yield (grain + straw) of rice and wheat increased by 3.64 Mg C ha−1 (71%) and 2.99 g C ha−1 (145%), respectively with NPK + FYM application, compared with UF control. Average across the treatments, total C yield (grain + straw) of rice was 82.6% higher than the wheat. CER of rice and wheat increased significantly with NPK + FYM application than the application of FYM alone (Table 5). The highest CER for rice (2.22) and wheat (1.41) under NPK + FYM, compared with other treatments indicates that NPK + FYM application had the highest potential to fix C per unit of the amount of C loss as GHG emission.

Table 5 Effect of nutrient management practices on grain and straw C yield and carbon efficiency ratio (CER) of rice and wheat grown in a sequence in northwestern India
Full size table

3.5 Economic Analysis of Rice–Wheat Cropping System

The effect of different nutrient management practices on economic of rice and wheat production was assessed through GR, NR, benefit–cost (B/C) ratio, GRAFC, and the economic efficiency (Table 6, Fig. 5). The highest gross returns from rice (GRR) and net returns from rice (NRR) were obtained in NPK treatment, followed by NPK + FYM and were the lowest in UF control. The B/CR in NPK treatment was higher by 1.57, compared with UF control. GRAFCR were also higher in NPK, compared with other treatments. The gross returns from wheat (GRW) and net returns from wheat (NRW) were, however, higher in NPK + FYM treatment, compared with NPK alone. B/CW improved by 1.97 with NPK + FYM application over the NPK alone. GRAFCW were higher in N + FYM and NPK + FYM treatments. Over an annual cropping system, GRAFC with NPK + FYM were increased by 79 US$ ha−1, compared with NPK alone. Average across the treatments, economic efficiency of rice and wheat was 8.3 US$ ha−1 day−1 and 6.0 US$ ha−1 day−1, respectively (Fig. 5). Economic efficiency of rice and wheat increased by 3.8 US$ US$ ha−1 day−1 and 5.2 US$ ha−1 day−1, respectively, with NPK + FYM application, compared with UF control. Economic efficiency in UF control and PK treatments were more or less nearly the same, and increased considerably with N application (N, NP, NK, and NPK). The NEEB varied between 1338 and 2512 US$ ha−1 in response to the effect of different treatments on grain yield of rice–wheat cropping system. NEEB exhibited a significant linear relationship with GWP of rice–wheat system (Fig. 6), which was best described by a linear function (Eq. 40).

$$ \mathrm{NEEB}\ \left(\mathrm{US}\$\ {\mathrm{ha}}^{-1}\right)=402.6\ \mathrm{GWP}\ \left(\mathrm{Mg}\ {\mathrm{CO}}_{2\mathrm{e}}\ {\mathrm{ha}}^{-1}\right)-2854,{R}^2=0.973\ast \ast, p<0.01 $$
(40)
Table 6 Effect of nutrient management practices on economic indices viz. cost of cash inputs (US$ ha−1), gross returns (US$ ha−1), net returns (US$ ha−1), benefit–cost ratio, and gross returns above fertilizer costs (US$ ha−1) for rice-wheat cropping system in northwestern India
Full size table
Fig. 5
figure 5

Effect of different nutrient management practices on economic efficiency of rice and wheat in northwestern India. Line bars indicate standard error from mean

Full size image
Fig. 6
figure 6

Relationship between global warming potential (GWP) and net ecosystem economic budget (NEEB) for rice-wheat cropping system in northwestern India. Line bars indicate standard error from mean

Full size image

3.6 Energy Indices of Rice–Wheat Cropping System

Impact of different nutrient management practices on energy input–output indices for rice and wheat was assessed through NEG, EP, ES, EI, and ER (Table 7). The NEG was highest in NPK + FYM (467 GJ ha−1) and lowest in UF control (252 GJha−1). Omission of fertilizer N (PK) in rice–wheat cropping did not result in NEG compared with UF control. On the other hand, N application (N, NP, NK, NPK) significantly increased the NEG between 172 and 195 GJ ha−1 (by ~ 68–77%) over UF control. Alone application of NPK and FYM increased the NEG by ~ 4.5% and 69.8%, respectively, than NPK + FYM. Average across the treatments, NEG through rice cultivation was ~ 68.7% higher than in wheat. A significantly higher EP and the lower SE and ER in PK compared with other treatments including fertilizer N was due to increase in productivity of rice and wheat. Higher ES and lower EP values in manure applied treatments (N + FYM and NPK + FYM), compared with the application of inorganic fertilizers was observed. NPK + FYM application increased EI by 7.3 MJ US$−1 and 27.6 MJ US$−1, respectively, over NPK and FYM alone.

Table 7 Energy indices for grain + straw produced in rice-wheat cropping system under different nutrient management treatments
Full size table

4 Discussion

After 5 cycles of rice–wheat cropping with manure and fertilizer application, nutrient availability increased significantly. The accumulation of NO3−-N and NH4+-N concentrations with manure application could be ascribed to N addition through FYM and increase in N mineralization (Dinesh and Dubey 1999) because of favorable soil conditions for microbial activity (Zhang et al. 2009). The combined application of chemical fertilizers and manures (NPK + FYM and N + FYM) resulted in higher concentration of mineralizable N, compared with their alone application (N, NPK, and FYM) (Mohanty et al. 2013). The amount of N released from FYM and not immobilized by soil microflora was 2 times higher in the soil that had been fertilized with 60 Mg ha−1 year−1 of FYM as compared to the soils without FYM application (Nett et al. 2010). About 47–62% of manure N gets mineralized within initial 28 days of incubation (Mohanty et al. 2013). The increase in potentially mineralizable N following long-term manure application has been ascribed to increase in easily degradable proportion of soil organic matter (Whalen et al. 2001).

The integrated use of manures and chemical fertilizers results in enhanced nutrient use efficiency and sustainability of the agricultural production (Benbi and Biswas 1997). The authors reported greater accumulation of P in soils with NPK + FYM application compared with NPK alone. The organic manures are rich source of organic P which influences P availability due to conversion of insoluble Ca, Fe, and Al phosphates to soluble forms through the formation of chelates and decomposition products of organic matter (Stevenson 1982). Phosphorus accumulation with FYM application was attributed to P addition through manure (Toor and Bahl 1997), increased P mineralization (Singh et al. 2010), and decreased P sorption in soils (Sui and Thompson 2000; Singh and Singh 2011). The synergistic effect of fertilizer P conjointly used with manures on P availability was more than sum of the increase from either applied individually (Toor and Bahl 1997). Manure application along with inorganic P in soils tends to shift the P solubility towards super-saturation with respect to dicalcium phosphate dihydrate (DCPD), indicating enhanced P availability as a consequence of lowering of phosphate potential (Singh et al. 2010). Increased K concentration with the application of fertilizers and organic manures has also been reported (Naidu et al. 2009; Venkatakrishnan and Ravichandran 2012).

A significant increase in grain yield of rice with application of N and P (in NP treatment), compared with N alone could be ascribed to high P sorption in soils of the region (Toor and Bahl 1997; Singh et al. 2010). In slightly alkaline and calcareous semiarid subtropical soils, most of inorganic P exists as calcium-bound P (Pasricha et al. 2002) which led to lesser P availability to plants. In order to maintain optimum soil P fertility, P application in excess of crop uptake by 10–30 kg P ha−1 year−1 is required to compensate for P sorption and immobilization (Aarts et al. 2000). Potassium application together with N only increased the grain yield of rice, but had non-significant effect on wheat grain yield. Earlier studies conducted in the study region showed that K application gave response of 280–290 and 160–240 kg grain ha−1 in wheat and rice, respectively (Randhawa and Tandon 1982; Meelu et al. 1995). In this study, application of NPK and manure (NPK + FYM treatment) magnified the rice and wheat grain yield response, which implies beneficial effect of manure in terms of enhanced nutrient availability. Shang et al. (2011) also reported that application of organic fertilizers in combination with adequate nutrients improved crop yield.

Nutrient uptake increased in proportion to the crops’ response to fertilizer application in rice–wheat cropping system. The nutrient use efficiency was assessed through yield increase due to nutrient application compared with UF control through AE and the apparent RE calculated as percentage of nutrient recovered in above-ground biomass at harvest. A significantly higher AEN in rice and wheat in NPK treatment, compared with other treatments showed the influence of balanced use of fertilizers. The increase (in AEN) of lower magnitude in NPK + FYM was due a small change in rice and wheat yield due to additional N input through FYM, compared with NPK alone. The AEN of 15.3–21.4 kg (kg N)−1 in rice observed in this study was in range 16–25 kg (kg N)−1 to that reported earlier (Witt et al. 2007). In lowland rice, AEN of 15–18 kg kg−1 at farmers’ fields in the Philippines has been reported (Cassman et al. 1996). Average across the treatments, REN was 24.5% in rice and 20.2% in wheat in this study. The average REN of 30% for irrigated rice in Asia has been reported (Dobermann and Fairhurst 2000).

The minimum and maximum values of PEN, PEP, and PEK in rice were 66.7 to 92.8 kg N kg grains per kg plant N, 152 to 366 kg grains per kg plant P, and 79 to 357 kg grains per kg plant K, respectively. These values are lower than the range reported earlier by Chuan et al. (2013). The optimal PEN of 68 kg kg−1 for irrigated rice reported by Witt et al. (1999) implies that most fertilizer treatments in the present study had optimal PEN in rice and there is no intrinsic obstruction in achieving high AEN in rice. A quantum leap in AEN is possible in irrigated rice and wheat with balanced fertilizer application (NPK), compared with imbalanced use of fertilizers. In the Indo-Gangetic Plains, Nambiar (1994) observed increased N use efficiency with NPK application in several long-term experiments on rice–wheat cropping system over the use of fertilizer N alone. In a 22-year long-term experiment on maize–wheat–cowpea cropping system in northwestern India, Benbi and Biswas (1997) reported that REN in wheat was directly related to the balanced (NPK) application of fertilizers.

In this study, RIUEN, RIUEP, and RIUEK in rice were lower than reported earlier (Pathak et al. 2003; Liu et al. 2006; Buresh et al. 2010). Using a dataset of more than 8000 field observations in the QUEFTS (Quantitative Evaluation of the Fertility of Tropical Soils) model, Buresh et al. (2010) predicted RIUE at 60–70% of rice yield potential corresponded to plant accumulation of 14.6 kg N, 2.7 kg P, and 15.9 kg K per Mg of rice grain that yielded N/P/K ratio of 5.40:1:5.89. The N/P/K ratio for rice in this study was however on the lower side (2.04:1:1.82) because of lower RIUEN and RIUEK and higher RIUEP. Witt et al. (1999) reported that 25 kg K + 9 kg P ha−1 are required to increase the nutrient restricted yield by 1 Mg rice grain yield ha−1. A linear increase in grain yield was predicted using QUEFTS model with nutrient uptake in balanced amount of 14.7 kg N, 2.6 kg P, and 14.5 kg K per Mg of rice grain until yield reached 60–70% of the yield potential. The corresponding PEs of 68 kg grain kg−1 N, 385 kg grain kg−1 P, and 69 kg grain kg−1 K with balanced nutrition have been reported (Witt et al. 1999). At 60–70% potential yield, Chuan et al. (2013) reported that 26.4 kg N, 6.5 kg P, and 22.0 kg K would be required to produce 1 Mg wheat grains with corresponding N/P/K ratio of 4.06:1:3.38. The N/P/K ratio for wheat in the present study was comparatively lower (2.07:1:1.76), possibly because of lower RIUEN, RIUEP, and RIUEK to produce 1 Mg wheat grains. Witt and Dobermann (2004) reported that in irrigated rice across Asia, REP varied between 22 and 35% and REK from 35 and 66% between median and 75% percentile data range.

The CO2, CH4, and N2O are the most important GHGs that contribute to global warming. The net emission of CO2e from different farming activities could be partly offset by altering management practices that can enhance soil organic matter content and/or decrease their emission (Mosier et al. 2006; Singh et al. 2020; Singh and Benbi 2020). However, the release of CO2 during the manufacturing and application of N fertilizer to crops can counteract these mitigation efforts (West TO and Marland 2002). The net GWP in our study (10.4–12.3 Mg CO2e ha−1 year−1) was higher than 6.7–9.7 Mg CO2e ha−1 year−1 as reported earlier for rice–wheat cropping system (Ma et al. 2013). These estimates were within the range of 8.7–11.6 Mg CO2e ha−1 year−1 reported by Xia et al. (2016) in rice–wheat cropping system in Chinese croplands, with CH4 emission being the dominant component of GWP in all treatments. In a rice–wheat cropping system, Bhatia et al. (2005) reported a range of 11.8–17.9 Mg CO2e ha−1 for different nutrient management regimes. Singh and Benbi (2020) reported that the GWP for rice fields that were intermittently flooded with single and multiple aerations was lower by ~ 18.1 and 27.6%, respectively, than the rice fields remained continuously flooded throughout the growing season. Xia et al. (2016) reported that substitution of 100% inorganic N by organic N led to significant increase in GWP due to increase in CH4 emission. These results corroborate the earlier research findings of Singh and Benbi (2020) who reported that GHGI for rice under continuous flooded conditions was significantly higher compared with the maize-based ecosystem with aerobic soil conditions throughout the year.

Albeit of considerably higher GWP for NPK + FYM treatment compared with UF control, the lowest GHGI (3.69 kg CO2e kg−1 grain yield) was obtained under NPK + FYM treatment. It implies less consumption of CO2e per unit of grain produced with conjoint application of manure and fertilizers. The GHGI is affected both by N input and source, due to the effect on grain yield and nutrient use efficiency (Pan et al. 2009; Shang et al. 2011; Singh and Benbi 2020). Liang et al. (2013) also suggested that application of organic fertilizer increased GHGI in rice ecosystem. These results are consistent with the findings of Burney et al. (2010) who reported that higher grain yield help offsets the GHG emission. The increased GWP of a rice–wheat cropping system due to conjoint application of NPK + FYM compared with the application of NPK has been attributed to enhanced CO2 emissions due to higher substrate availability for oxidation (Bhatia et al. 2010). Although, manure and fertilizer application increased the GHG emission and the associated GWP of a rice–wheat cropping system, it increased the total soil organic C (TOC) content in this cropping system, which has been published earlier (Singh and Benbi 2018). Six et al. (2004) also reported that GHG emission substantially offsets TOC increase and vice versa, and there is a trade-off among GHG emission in crop production. The nutrient management strategies which reduce fertilizer application rates without influencing crop yields could inevitably lower the GHG emissions (Mosier et al. 2006; Singh and Benbi 2020).

The GRAFC for rice was higher for NPK compared with NPK + FYM because of non-significant yield differences among the two treatments, rather NPK + FYM had extra expenditure of 22 US$ ha−1 incurred on manure application. In contrast, NPK + FYM treatment yielded higher GRAFC in wheat than NPK alone, possibly because of significant increase in wheat grain yield in manure applied treatment with similar CCI of 265 US$ ha−1. It was further confirmed from the effect of two treatments on the economic efficiency of rice and wheat with significant differences for wheat only. Manure application (in NPK + FYM treatment) also increased the NEG in wheat to a higher extent, compared with rice which implied extra energy input associated with its application before rice establishment. The non-significant difference for NEG under NPK and NPK + FYM treatments was because of a non-significant change in the rice grain yield. A significantly higher EP and ER for PK treatment, compared with other treatments was ascribed omission of N which is the major fertilizer-related energy input in rice–wheat cropping system. Southwell and Rothwell (1997) reported that energy requirement of fertilizer N was much higher compared with P and K. It was ascribed to higher energy equivalents of 60.6 MJ (kg N)−1, which were 5.5 and 9.0 times higher compared with 11.1 MJ (kg P2O5)−1 and 6.7 MJ (kg K2O)−1 (Singh et al. 2002; Demircan et al. 2006; Esengun et al. 2007). High energy equivalent of N fertilizers is due to the energy-intensive Haber–Bosch process of its production involving natural gas (Nemecek and Erzinger 2005). Relative amount of energy input in different cropping systems comprised 44–54% for chemical fertilizers (Yadav et al. 2017). In a rice–wheat cropping system, Khan et al. (2009) reported chemical fertilizers comprise 43% for rice and 47% of the total input energy used for wheat. In the study region, Singh et al. (2019a) reported that of the total input energy of 52.4 ± 1.3 GJ ha−1, ~ 24.7% was shared by chemical fertilizers. However, in wheat, chemical fertilizers comprised ~ 45% of total energy input 25.6 ± 0.5 GJ ha−1 (Singh et al. 2019b). The higher energy input resulted in lower ER and EP (Billore et al. 1994; Singh et al. 2020). The management practices producing greater differences among EO and EI leads to higher NEG (Singh and Benbi 2020), while higher ER implies potential to be used as energy management indicator for environmental and economic sustainability (Tzanakakis et al. 2012; Pishgar-Komleh et al. 2012).

5 Conclusions

The results of the study highlights significance of balanced use of chemical fertilizers (NPK) along with manure application in terms of increased crop productivity, net energy gain (NEG), C budget and economics of rice–wheat cropping system due to increased nutrient availability and nutrient use efficiency. Although, global warming potential (GWP) was highest in NPK + FYM treatment, yet the yield-scaled GWP estimated as greenhouse gas intensity (GHGI) decreased by ~ 35.8%, compared with the application of FYM alone. The highest C efficiency ratio (CER) for NPK + FYM indicates the highest potential to fix C per unit C loss as greenhouse gases (GHGs) emission in rice–wheat cropping system. The conjoint application of NPK + FYM increases energy intensiveness (EI) and NEG of rice–wheat cropping system, compared with their separate application. The net ecosystem economic budget (NEEB) increased with application of chemical fertilizers conjointly with manure (N + FYM and NPK + FYM). The study showed that NPK + FYM application is the most appropriate nutrient management strategy for realizing higher economic and C yields and nutrient and energy use efficiency (ER) with considerable potential to mitigate GHG emission in rice–wheat cropping system.