Introduction

To meet the challenges posed by climate change and other environmental adversities, many tools and indicators have been developed to assess the environmental impacts of various systems (Finnveden et al. 2009; Nie et al. 2011). Life cycle assessment (LCA) is one of these major tools that can evaluate environmental burden posed by a product or an activity (Ayres and Ayres 2002; Haes 2002; Rebitzer et al. 2004; Finnveden et al. 2009).

LCA is a widely accepted impact assessment tool in the field of industrial ecology (Haes 2002). It can methodologically determine the environmental impacts attributable to each life phases of a product’s entire life cycle (Consoli et al. 1993; Miettinen and Hämäläinen 1997; Rebitzer et al. 2004). Therefore, LCA is instrumental in obtaining a comprehensive picture of a product’s environmental performance during its life span, to enable informed decision and policy making (Miettinen and Hämäläinen 1997). Finnveden and his colleagues (2009) also reported that due to LCA’s unique methodology, it avoids burden-transfer from one phase of life cycle to another, or from one environmental impact to another. This comprehensive scope of LCA makes it a very robust tool for environmental impact assessment and assist in discovering sustainable options. Life cycle assessment has also seemingly become the gold standard for emerging environmental certification and labeling programs (Pelletier 2014). This is a testament that LCA has the capacity to bridge consumers and producers of the world, integrating environmental considerations. Therefore LCA plays an important role for authorities, individuals and all industries in achieving a wide range of benefits (Miettinen and Hämäläinen 1997; Roy et al. 2009).

The world population stands at 7.4 billion in 2016 (Population Reference Bureau 2016). To sustain them all, we have taken over almost half of the world’s land mass and turned it into our primary source of sustenance and livelihood (Owen 2005). Agricultural food production is putting a huge amount of environmental burden (Tilman et al. 2001; Garnett 2011). The result of agriculture and the intensity of cultivation practice leave a lasting and irreversible impact on the environment (Önder et al. 2011). Agricultural practices consequently contributes heavily to greenhouse gas emissions, soil fertility, biodiversity loss and release of toxic substances in the ecosystem among other pollutants (McMichael et al. 2007). 10–12% of the global heating reportedly comes from agricultural activities (Mogensen et al. 2015; Mosleh et al. 2015). Besides climate change, agricultural practices contribute to an array of environmental impact categories (Önder et al. 2011).

Hence there is an urgency to evaluate both our production and consumption pattern; before the environmental impacts associated with food production exceeds natural boundaries. Soussana (2014) reports that using life cycle perspective may lead to breakthroughs in the sustainability assessment of food systems; it is therefore unsurprising that literature of LCA to food systems has been thriving recently (e.g. Notarnicola et al. 2012, 2017; Nemecek et al. 2016; Goossens et al. 2017). There is an abundance of literature related to LCA on agricultural crops such as potato (Mattsson and Wallén 2003; Veisi et al. 2017), tomato (Andersson et al. 1998; Nienhuis and de Vreede 1996; Van Woerden 2001; Anton et al. 2005; Roy et al. 2008; Torrellas et al. 2012), sugar beet (Bennett et al. 2004), wheat (Brentrup et al. 2004; Charles et al. 2006), dry pea and lentil crops (MacWilliam et al. 2014) and apple (i Canals et al. 2006; Goossens et al. 2017). However, despite being the fundamental source of sustenance and environmentally intensive, we are still unaware of full phase LCA case study of rice, especially in developing countries. Our study is an attempt to fill this gap.

Among all agricultural commodities, rice is the most important (Breiling et al. 1999; Khush 2005; Roy et al. 2009). Rice is the cardinal diet of over half the world’s population (FAO 2011; Muthayya et al. 2014). Currently grown in more than 100 countries and all continents except Antarctica, rice is integral for world’s economy and food security. With 90% of world’s production (Muthayya et al. 2012), Asia is the world’s largest rice-producing and rice-consuming region and is also becoming an increasing food staple throughout Africa (de Miranda et al. 2015).

We are aware about some notable LCA studies on rice in developed countries viz. Italy (Blengini and Busto 2009), Japan (Breiling et al. 1999; Harada et al. 2007; Hokazono et al. 2009; Hatcho et al. 2012; Yoshikawa et al. 2012; Hokazono and Hayashi 2015), USA (Brodt et al. 2014). However, LCA as a tool is largely unexplored in developing countries. World rice consumption estimate for 2013–2015 was reported as 490,804,000 tonnes per year, where developing countries alone consume 471,919,000 tonnes (96%) (OECD/FAO 2016). The volume of consumption in this region alone provides compelling ground that LCA should be promoted in developing countries to become standard practice in agriculture sector for sustainable and continuous generation of crops.

Bangladesh is the epitome of rice based nation, as it stands fourth in top rice producing countries of the world, as well as holds the second place for rice consumption per year per capita (169.5 kg per capita), only second to Vietnam (OECD/FAO 2016). Agriculture makes up 15.1% of Bangladesh’s GDP (CIA 2017). The importance of agriculture in the country also lies in the fact that almost half of the population earns their livelihood in this sector, as per the world bank database (The World Bank 2010). Despite being a small country, Bangladesh produces a wide range of crops viz. rice, wheat, maize, pulses, tobacco, cereal, potato, jute, oil seeds, spices, chilies, onion, tea, drugs and narcotics, vegetables, tomato, fruits and sugar crops (BBS 2012) (Fig. 1). Although Bangladesh is the fourth highest producer of rice, it is the smallest country to have such a high yield of production. An incredible 80% of the total country area is cultivated for rice harvest (CIA 2017; USDA 2017).

Fig. 1
figure 1

(Source: FAO 2017)

Annual production of prominent agri-crops in Bangladesh from 2004 to 2014

Full size image

However, the agriculture scene of the country is tethered with crude agricultural practice, use of conventional, outdated machines (BBS 2012) and obsolete pesticides (Parveen and Nakagoshi 2001). Bangladesh also tend to have intense cropping practices. With triple cropping system in a season, the agricultural inputs can environmentally be burdening. Especially to meet food demand, agriculture has been intensified through use of chemicals, and high yielding variety of crops; a stark contrast of cultivation practice as done by developed parts of the world (Uddin and Takeya 2006). Organic farming is also not prominent in the culture of Bangladesh (Hoque 2012). Although the nation is well suited for agriculture in terms of climate and soil type, high production of rice typically equates to tremendous consumption of water through irrigation (Chowdhury 2010). Especially in terms of application of commercial fertilizers, Bangladesh is within top 20 most fertilizer using country in the world (FAO 2017). This is a staggering number, considering the small size of the country and high fertile soil. According to FAO statistics (FAO 2017), Bangladesh is only behind China in terms of volume of fertilizers used in agriculture per hectare area (Fig. 2).

Fig. 2
figure 2

(Source: FAO 2017)

Fertilizer use among highest rice producers

Full size image

Rice paddy production in Bangladesh have increased gradually over the years, and is not likely to slow down due to constantly increasing food demand. Bangladesh also export surplus rice variety overseas, which contribute to the country’s welfare (BRRI and BBS 2016). Given that rice is a highly produced crop in South Asia, there is dire need to evaluate its environmental performance at each step of its life. The demand for food production is projected to increase within 2050, by a large margin. If this trend continues, the environmental stress will increase due to increase in rice production to meet future demand. Hence, conducting LCA on rice is not of national interest, but urgent for global context as well.

According to FAO (2017), over 30% of carbon dioxide equivalent emission in Bangladesh are contributed from rice cultivation. Through life cycle assessment, it will be possible to identify environmental hotspots/weakspots in its life cycle at every phase of production, which can then be used to prioritize stages of the production chain needing improvement to mitigate harmful emissions. By conveying the results of an LCA, it may be possible to introduce new policies and put a break to the vicious cycle of intense cropping practices. While it is important that rice production increases with demand, the production process must be streamlined as such, that it performs optimally not only financially but from an environmental standpoint. It is urgent that we look for opportunities to find room for improvement, to meet demands and improve yield without compromising environmental integrity or public health.

Use of LCA in industries of Bangladesh is also unheard of. However, the institutional set-up of the society can be rapidly changed through popularizing the use of LCA in the industries of Bangladesh. Once the appeal of benefits LCA can bring in optimizing resources is understood, there will be an increased enthusiasm in keeping comprehensive and up-to-date directories for transparency between producers and consumers. In light of that, the LCA on rice paddy production is the first attempt at a gate to gate study in this region. As a consequence of the significant contribution of the results in the case study, use of LCA will be popularized, so that, in the long run, it becomes statutory to use the comprehensive and holistic tool to optimize processes in an activity or production operation in Bangladesh or any other developing countries of the world. Toxic substances and environment stressing activities are probably a larger problem in developing countries like Bangladesh, and should get more attention from LCA practitioners (Haes 2004).

LCA on rice in Bangladesh has been done on one occasion alone, as author Roy et al. (2007) conducted a life cycle assessment of CO2 emission from parboiled rice production and compared it with untreated system of rice production. However, the mentioned author’s case study did not include LCA during the agriculture (cultivation) phase of rice production; which is the production of paddy rice in the field. Hence this study adds a new nexus by establishing the total impact that is enclosed in the generation of paddy rice in a typical scenario, to provide an average scenario of impacts on the environment. This will provide environmentally conscious policy makers, producers and consumers with comprehensible information to make informed steps, with the help of LCA—to explore different options and greener alternatives.

Methodology

Overview of the methods

The life cycle assessment methodology has been used to evaluate the environmental profile of general rice paddy farming. Standard guidelines for life cycle assessment and more specifically, life cycle assessment of agricultural crops were followed (Tillman and Baumann 2004; Nemecek et al. 2014). Goal and scope definition is aimed at establishing the study objectives, functional unit (FU), and system boundaries and data sources.

Life cycle inventory (LCI) stage details out all the environmental inputs and outputs at each stage of the case study within the defined system boundary. Life cycle inventory is the backbone of an LCA (Jiménez-González et al. 2000). The objective of the inventory is to create a model of the product or activity identified during defining the system boundary in the goal and scope definition stage. Primary data collection for this stage was mined from conducting field survey from two different farms in Bangladesh, in Meherpur and Baliakandi, where about 20 farmers were interviewed. As information from farmers can result in varying information, as cultivation practice can differ from region to region, also depends on farmer’s individual financial capacity and resource availability, private communication with rice research scientific officer at Bangladesh Rice Research Institute (BRRI) was also established, for reliable data, that would serve the purpose of modelling the LCA according to typical cultivation practice. Data were also taken from literature study and cited accordingly.

Life cycle impact assessment quantifies the relative importance of all the environmental burdens identified in the LCI by analyzing their influence on selected environmental categories. Life cycle impact assessment is composed of classification and characterization that converts LCI result into an indicator representative of each impact category. For this analysis, the ReCiPeFootnote 1 Midpoint (H) methodology was employed (Huijbregts et al. 2017). ReCiPe uses 18 midpoint indicators and three endpoint indicators. Endpoint was not considered in this study. The midpoint indicators are in the table below (Table 1).

Table 1 List of ReCiPe midpoint indicators
Full size table

LCA of rice paddy in Bangladesh (baseline scenario)

Rice paddy production in Bangladesh has technically similar practice as other countries in the world, but differing in intensity, use of technology and cultivar variety. It was still important to establish the steps of paddy cultivation, by interviewing farmers in the field, personally communicating with Bangladesh Rice Research Institute and comparing literature studies. Details of data gathered and sources are discussed in subsequent sections.

The LCA model describes a typical farm in Bangladesh. Average quantity of inputs applied (fertilizers, pesticides, etc.) are taken into consideration. Average yield harvested since 1995 to 2015 roughly corresponds to 4 t/ha (FAO 2017). The cultivar is not specified as the case study is based on average grain yields over a long period; this exclusion is similar to authors who have studied the life cycle assessment of milled rice in Italy (Blengini and Busto 2009).

System boundaries and functional unit (goal and scope)

For the present analysis, the LCA model was carried out as “gate to gate” in the cultivation field (Fig. 3) and not a complete “cradle to grave”, the activity starts with soil cultivation after the harvest of the previous crop. The activity ends with the harvest of rice grains and subsequent burning of crop residues.

Fig. 3
figure 3

System boundary for this study

Full size image

The system boundary have been divided into foreground system and background system. Foreground system is all the agricultural activity and inputs that are made after land is prepared to grow the rice paddy after harvesting of previous crop is finished. The agricultural input that are taken into consideration include all machine operations, corresponding infrastructure, and fuel use. Machine operations are: rotary tillage, the application of pesticides and fertilizers using machines, irrigation pumps and harvesters. Amount of pesticides, fertilizers and rice seed are accounted for. These are the foreground system, or the immediate agricultural activities within system boundary of the gate to gate study. The foreground activity produces environmental load as shown in the figure (Fig. 3), and rice paddy as the final outcome. The life cycle inventory created for the production of rice paddy is for foreground activity and are based on primary and secondary data.

The LCA model also includes background system, which is all the upstream activities that are required to complete the foreground activity. Namely, the production and manufacturing of plant management applications, machines, diesel and their respective transportation to the farm. The upstream activity also puts environmental burden, and the impact is accounted for during the life cycle impact assessment phase of this study. The data for background system has been obtained from ecoinvent database version 3.3, as discussed in later chapters.

Rice seedlings are transplanted manually by throwing the seedlings in the standing water. Paddy rice is grown under submerged conditions (50 cm standing water for approximately 150 days, assuming non-flooded conditions 1 week prior to harvest). Further, direct field emissions and land use change are included. The inputs of seeds, irrigation water, fertilizers and pesticides are considered.

Rice straw was excluded from the agriculture phase, and no allocation criteria had to be employed in this regard as most of the rice straw is burned in site and their emission is considered neutral (Williams et al. 2006). Hence byproduct, straw, has been omitted from the system boundary. As the main function of the system under the study is rice paddy produced in the field, the functional unit selected is 1 kg of rice paddy produced at field. Delivery to the farm or supermarket is also outside the set system boundary.

Data sources

The data for the life cycle inventory were obtained from different sources. On site records of what fertilizers and pesticides are applied as well as average working hours of machineries and their fuel consumption were obtained through interviews with farmers of different farm locations aforementioned. Personal communication with rice agriculture scientific research officer as well as specific literature (Basak 2010) provided typical amount of fertilizers that are applied during agricultural process. The machineries used were mostly retrieved from established factsheet produced by BBS (2012). Information of pesticides were taken from literature data by authors Dasgupta (2003) and Parveen and Nakagoshi (2001), and cited accordingly. Data for greenhouse emissions (methane), were taken from literature study (Khan and Saleh 2015), which are in par with agricultural LCA methodology and guidelines (Nemecek et al. 2014). Water use figure was used from literature (Chowdhury 2010), which stated average water usage; from which water consumption and emission were calculated using the guidelines (Nemecek et al. 2014). For the inputs where there were limited supporting data to conduct calculation, proxy values were used. The proxy values for ammonia, field emissions were taken from ecoinvent databaseFootnote 2, under the life cycle inventory of rice paddy production in India. Background data for all upstream activities of relevant inputs were taken from ecoinvent database 3.3 (Table 2).

Table 2 Paddy rice production phases and data sources
Full size table

Inventory analysis

This section will explain what data were collected, and how they were adapted to the LCA model to carry out the analysis using openLCA version 1.6 and ReCiPe’s midpoint methodology (Huijbregts et al. 2017). Assumptions made and proxy data taken are mentioned here.

Field operation

Field operation include ploughing, fertilizing, irrigating with diesel powered pumps, sowing, application of plant protection products and harvesting. The types of machineries used were retrieved from Bangladesh Rice Research Institute’s factsheet and Bangladesh Rice Agriculture Statistics (BBS 2012). The entries were made using ecoinvent database version 3.3, which included the manufacturing data of each machineries. The values entered were adapted from considering average working hours and fuel consumption, obtained from interviewing farmers in the field survey conducted in Meherpur and Baliakandi farms in Bangladesh (Table 3).

Table 3 LCI of rice paddy production: model entries for field operation and manufacturing
Full size table

Fertilizers

The use of different types and amount of fertilizer can vary to a great extent from one farm to another, especially of different regions of the country. Application also depends on the rice species and soil. The average use of fertilizers was estimated using a factsheet provided by senior scientific officer at Bangladesh Rice Research Institute as well as literature data (Basak 2010). The data was substantially similar when cross checked by personally communicating with rice farmers. There are naturally variation in quantity of application, from farm to farm, however, the variations were not considered typical and hence only the general prescribed amount is considered. The manufacturing processes (background system) were taken from ecoinvent database version 3.3 (Table 4).

Table 4 Model entries for fertilizers and sources
Full size table

Pesticide

As the use of pesticide vary from farm to farm in Bangladesh, data from two sources were used to model the LCA (Parveen and Nakagoshi 2001; Dasgupta 2003). Bangladeshi farmers have been recorded to use several classes of pesticides, herbicides and fungicides over the years, hence the most widely used pesticides based on interviewing farmers, were selected. According to the active ingredients, inventory data from ecoinvent were appropriately selected for the model (Table 5).

Table 5 Model entries for pesticides
Full size table

Irrigation

Irrigation in Bangladesh can be intensive. According to literature data, average irrigation requirement is 11,500 m3/ha for one harvest (Chowdhury 2010). Water consumption was calculated using blue water consumption method, stated in the guideline for agricultural LCA (Nemecek et al. 2014). Green water (precipitation, moist) etc. were not considered as they do not have any tax on the environment. Water emissions were based on the equations provided in the same guidelines (Table 6).

Table 6 Irrigation input and water emission
Full size table

Emissions from rice paddy field

Field emissions include direct air emissions of methane, nitrous oxide and ammonia, as well as emissions of phosphorous and nitrates to water. Anaerobic decomposition of organic matter and the consequent methane production are caused by intense water management practices because rice production hinges on long submersion times. This is however magnified with the use of fertilizers (Watanabe et al. 1995; Butterbach-Bahl et al. 1997; Mitra et al. 1999; Dan et al. 2001; Majumdar 2003; Harada et al. 2007; Arunrat and Pumijumnong 2017).

Carbon dioxide emissions were calculated using the instructions set in the world food life cycle database guideline, which is calculated on the basis of Urea applied as input during cultivation process. Methane emission were directly adopted from case study conducted in Bangladesh, where greenhouse gas emission in rice paddy field was calculated using IPCC guidelines (Khan and Saleh 2015). Since this methodology is also prescribed in the WFLDB-guidelines, which is used to model LCA of agricultural crops, the result of methane emission from paddy field was directly usable.

Because of lack of literature data and case studies, values for nitrogen dioxide and ammonia emission for the LCA model were adapted from ecoinvent database. Emissions from India’s rice paddy production dataset was used as proxy, given the regional similarities, it served as the best proxy for this case study’s purpose (Table 7).

Table 7 Emission to air from rice paddy field; data and sources
Full size table

Phosphorous releases

Phosphorous releases were adapted from dataset of India, retrieved from ecoinvent database to serve as proxy due to lack of literature studies and cases covered in Bangladesh. According to the guidelines to create the life cycle inventory for agricultural products, phosphorous emissions can be calculated using set of equations and variables, however, because of inadequate information to carry out the complete methodology, proxy data had to be used in this regard, from ecoinvent dataset of India’s rice paddy production (Table 8).

Table 8 Phosphorous emission to water
Full size table

Carbon uptake by crop from air

During growth of crops, the plants uptake carbon dioxide from air. This has to be taken into account for a precise analysis. This was calculated using the instructions set in WFLDB guidelines. Carbon content of biomass was assumed to be 47.5% (Table 9).

Table 9 Carbon uptake data entry
Full size table

Land transformation due to land use

Land transformation is a change from one type of land to another due to anthropogenic use. The amount of land transformed is the area required to produce one unit of the functional unit of a product. In the case of this study, the functional unit is 1 kg of rice produced in 1 ha area of land. Hence the entries for land use change were entered according to this guideline (Nemecek et al. 2014) (Table 10).

Table 10 Entries for land use change
Full size table

Seed production

According to the average data from the Bangladesh Rice Research Institute’s factsheet, farmers use varying amount of seed per hectare area, on the basis of the cultivar used. For high yielding cultivar variety, often lower quantity of seed is required. The amount also varies by region, as different farming fields have to apply different amount to obtain optimum yield. Personal communication with rice research senior officer revealed that farmers are typically prescribed to apply 40 kg of rice seed per hectare. Interviews with farmers that was carried out in field survey stage revealed that as a general practice, seeds more than 40 kg (up to 60 kg) is applied depending on farmer’s financial ability and resource availability. Hence an average of 50 kg/ha was considered as the typical seed input. Data for manufacturing of the seed are taken from ecoinvent database. The LCA model included use of electricity for drying, and storage purpose. The model did not include transportation as the seed are manufactured domestically, and do not have significant distance from the farm gates.

Results and discussions

The results presented in this chapter are based on a functional unit defined as production of 1 kg of rice paddy.

Life cycle impact assessment and interpretation

The impact assessment phase was carried out by analyzing the results of the inventory in order to calculate the impact indicators detailed in previous section. Table 11 shows the indicators and the impact result of the LCA model for rice paddy production in Bangladesh. Some significant impact indicators are talked about in details in following sub sections.

Table 11 ReCiPe impact indicators and impact result
Full size table

Climate change as GWP100

As it can be seen in the table, to produce 1 kg of paddy rice, including all the inputs within the system boundary, the GWP100 indicator shows a carbon dioxide equivalent emission of approximately 3.15 kg per functional unit. This value represents the greenhouse gas emission, which is an aggregation of indirect and direct greenhouse emissions leading up to the production of paddy rice at each steps of its life cycle, and not only the greenhouse gas emission that is emitted directly from the paddy field.

Kasmaprapruet et al. (2009) reported an almost similar global warming emission of 2.9 kg CO2eq per functional unit (per kg of milled rice produced after dehusking). However, background system was not considered as the upstream activities started from cultivation process. Moreover, the emission may be overestimated, as it is unclear if the authors accounted for allocation criteria for milled rice produced, and by product of husk produced as a result.

Case study in Italy revealed that delivery of 1 kg of exported white milled rice from the farm under study, contributed a carbon dioxide equivalent emission of 2.9 kg (Blengini and Busto 2009). But in this case study, producing 1 kg of paddy rice emits more than 3.15 kg CO2eq. The difference can be explained by mentioning the farm in Italy has non mechanized irrigation system, which has zero contribution to global warming potential; as there is no diesel combustion, or the need to use powered pumps. Their yield in 1 ha was also considered very high (7 t/ha). Due to which emission per functional unit is significantly lower. This makes a valid case in point, that to optimize the environmental load and production, maintaining high yield is necessary to obtain a low impact per functional unit generated.

Overall, the global warming impact during cultivation is largely due to methane emission from rice paddy field (29.29%) (Fig. 4). To reduce methane emissions from paddy fields, the options include using enhanced rice production technology such as minimizing the use of green manure and substituting pre-fermented compost from farm residues, adding nitrate or sulfate containing nitrogen fertilizer to suppress methane gas production or; change rice cultivation practices (Alting et al. 1997).

Photochemical oxidant formation

Energy consumption determines the impacts for photochemical oxidant formation. Non-methane volatile organic compounds (NMVOC) emission of magnitude 0.0123 kg per functional unit is the result (Table 11). The main source of NMVOC is from fuel combustion, particularly from field operation and road transport, according to emission inventory of NMVOC from anthropogenic sources, a case study conducted in China (Wei et al. 2008). In case of paddy rice production in Bangladesh, a contribution analysis reveals that irrigation system contributes 61.24% of photochemical oxidant formation, approximately 10% comes from field operation by machineries and 16% from paddy field emission (Fig. 4). Some NMVOCs damage the ozone layer in the upper atmosphere, and are also toxic for human health (European Environment Agency 2010). With the consequences in mind, we can prioritize on the basis of contribution from each subsystem, and improve the rice production system in Bangladesh by replacing traditional operational tools with environmentally friendlier options.

Terrestrial acidification as TAP100

Impact assessment reveals 0.02827 kg SO2eq is the resulting output per functional unit (Table 11). 47.3% of the impact is contributed from paddy field emission (Fig. 4). Some studies state that use of fertilizers type and timing of the application can determine the emissions of NH3 (affecting acidification) (i Canals et al. 2006). Hence to mitigate this impact, further research needs to be conducted for precise contributions of terrestrial acidification by varying the types of fertilizers and their timing, to reach a sustainable rice production practice.

Water depletion

Direct use of water appears to be particularly intense, almost 3 m3 of water per 1 kg of paddy rice produced is depleted; an indication that irrigation efficiency must be increased in order to optimize water depletion for rice agriculture. Currently, Bangladesh’s standard irrigation practice is flooding the surface in most cases. This has an efficiency factor of only 0.45 (Nemecek et al. 2014). Perhaps, by employing sprinkler irrigation system or drip channeling as the standard practice, the extracted water can be more efficiently applied to all range of agricultural crops. The predicted trend of such a change can result in much less requirement of direct use of water.

Marine eutrophication and freshwater eutrophication

Eutrophication is the impact on ecosystems from substances containing nitrogen or phosphorous. The consequences of which, is growth of algae or plants resulting the occurrence of situation without oxygen in the bottom strata due to increased algal growth (Alting et al. 1997). Paddy field emission has the most dominating contribution to marine eutrophication (80%) (Fig. 4), with a total magnitude of 0.01588 kg Neq (Table 11). Whilst irrigation system contributes the most to freshwater eutrophication (76%) with a total magnitude of 0.00126 kg Peq.

Ozone depletion

Ozone depletion is majorly contributed by production of pesticide, fertilizers, machine operations in the field and transportation. Irrigation system also makes up 34% of the total impact magnitude of 0.159 mg CFC-11eq. Blengini and Busto (2009) reported 0.10 mg CFC-11eq, although this magnitude was for every kg of milled rice delivered. It is clear that paddy rice production alone, is more environmentally burdening in Bangladesh.

Contribution analysis

A contribution analysis has been carried out with respect to the life cycle impact assessment result obtained for the production of 1 kg of paddy rice.

Figure 4 can hence be helpful to appraise the role and significance of the different subsystems that contribute to the life cycle impacts of paddy rice production. From this figure, it is possible to get a comprehensive and quick “at-a-glance” idea of the greatest contributor to different categories of impact indicators; on the basis of the inputs made in the inventory phase.

Global warming is mainly influenced by emissions from the paddy field (29.29%), which can be attributed to have been magnified by high usage of synthetic fertilizers and pesticides. Intense irrigation due to the amount of water withdrawal and machineries produced to facilitate the process means the highest percentage of contribution in global warming have come from irrigation (53.65%). Irrigation process also has great impact on three other indicators (FEP, PM, FD). This is unlike the irrigation system in Italy, which uses complete non-mechanized system called drip irrigation employing complex water canals and gravity to efficiently irrigate their rice paddy field. This neither consumes resources, nor puts an environmental burden (Blengini and Busto 2009). Their study stated that the irrigation subsystem has no contribution to global warming, ozone depletion, acidification or eutrophication. Our contribution analysis also reveals that significant contributions to climate change result from fertilizer use (10.98%).

As expected, water depletion is dominated by irrigation (99.54%). The remaining fractions being used by seed production, fertilizer production and field operation. Uses in packaging, and other inputs are almost negligible.

Paddy field emissions have the greatest impact on terrestrial acidification (47.3%). Highly affected by fertilizer emissions, the type of fertilizer used and the timing of the application determine the emissions of NH3. The importance of LCA is automatically highlighted here as it can be a great tool to assess the best crop management options in terms of environmental performance. From the paddy field, marine eutrophication (78.20%) as well as emission of particulate matter (11.48%) is also considerably high. Perhaps this result emphasizes the need for further reliable and site specific data, opening windows of opportunity for excellent research, to develop a more sustainable agricultural practice to mitigate the field emissions.

It should be stated that fertilizer production has an immense impact on ozone depletion, contributing the highest for this indicator (32.47%). Fertilizers also have significant contributions to GWP (10.98%) and freshwater eutrophication (11.78%).

Field operation’s contribution to each category indicator is noticeable (Fig. 4). As the primary source of their operation and maintenance is nonrenewable resources such as diesel, it is no surprise that they have remarkable contribution to photo chemical oxidant formation indicator (9.71%), ozone depletion (9.86%) and fossil depletion (4.26%). They also make up a noteworthy percentage of other impact categories.

From the figure (Fig. 4) it is evident that the “criminals” in the defined system boundary for the study are the irrigation process, manufacture of synthetic fertilizers and pesticides and emissions to air taking place in the rice paddy field (field emission) during the growth period; therefore, establishing the environmental hotspots in the system.

Fig. 4
figure 4

Contribution of subsystems to different impact categories of rice paddy productions

Full size image

Impact magnitude and the need to identify acceptable limits

The impact analysis highlights the great potential of LCA, which can account for a large number of parameters and calculate complex production systems where both natural and industrial processes exist (Blengini and Busto 2009). However, while there is no set standard limit of the impact magnitude of rice production for each impact categories, it is hard to identify if the impact values obtained can be considered acceptable or not. Upon comparing results with case studies done on different geographic region, it can only emphasize the need to improve on those benchmarks. However, further research need to be conducted to find the global average of impact magnitude to better understand the range of impact value, which can be considered acceptable, and the magnitude which are beyond acceptable limits.

Perhaps this study can shed lights on the need of such a standard framework to develop good agricultural practice. Introduction of such policy will ensure standard and sustainable cultivation practice is maintained, to keep all impact indicators involved in agricultural crop production within acceptable limits. LCA can pave the way for such sustainable development, as its ability to account for large number of parameters is evidence, that a framework can be developed around its methodology which will be both holistic and robust.

Limitations of conducting LCA in developing country

A large volume of data is required to carry out LCA (Finkbeiner et al. 2014). This makes it especially challenging to conduct an LCA case study in developing country like Bangladesh. Data mining in Bangladesh can be very difficult and time consuming. Low standards of bookkeeping, administrative coverage and social performance means databases or archives do not have the best form of maintenance. Specific local literature are also outdated, have questionable reliability, or are not technically sound to provide the robust information required to maintain precision in the assessment.

Conclusion

The study shows that usage of fertilizers, pesticides and non-renewable resources generate a significant share of impact across all categories. It is clear that application of different products for crop management and cultivation practice affects the LCA results, thus using petrol, diesel or electricity as an energy source magnifies the emission of CO2, NOx, SOx, heavy metal emission etc. Different fertilizers, pesticides have very different emission patterns due to their characteristics as well as different specific toxicities and manufacturing methods. Hence LCA may be used to guide the selection of these inputs with the aim of reducing environmental impacts. Greener technology such as drip irrigation system, solar irrigation system can be implemented at a large scale.

An analysis of improvement scenarios has shown that although mitigation solutions can be reached through an organic farming method, which does not involve application of commercial fertilizers such as urea, solutions cannot be restricted to single life steps or limited aspect. Since the consequences on the subsequent life phases could dramatically reduce or cancel out the improvements; such as reduction of yield. However, use of more environmentally benign pesticide and fertilizer compounds must be promoted.

Nevertheless, this LCA study attempted an objective environmental audit of the environmental system by quantifying emissions at a system level, in an effort to identify environmental hotspots of production. The original hypothesis was in fact correct, as the environmental hotspots are the irrigation phase and agricultural emission from paddy field. Communication of these results to producers in the form of improvement opportunities would be a useful addition to changing producer behavior and reducing environmental emissions. The producers of agricultural crops will benefit from this by being able to optimizing their yield with minimum environmental impact. The consequences of the change will improve food safety, improve public perception of the products and encourage environmentally friendly practice in all industries.

In this context, it is imperative to link the LCA results and the socio-economical needs of producers. This is crucial for acceptance by the agricultural sector, for example, if producers see LCA as a tool for optimizing their operation and increasing competitiveness by facilitating communication with the rest of the supply chain, this reaction will be from a much positive standpoint. However, communicating the LCA results remain a topic for further research.

Modelling the life cycle of paddy rice is both demanding and work intensive. It involves a large number of agricultural and industrial processes. Although improvements and further research are necessary, mainly studies pertaining to direct field emissions of nitrogen dioxide and ammonia in Bangladesh rice paddy production are of utmost importance. The present research has supplied quantitative results and information that might be useful in the future investigations.