https://orcid.org/0000-0002-5563-4276
Figures
Abstract
Efficient nitrogen (N) management is critical for sustaining high maize yields while minimizing environmental impacts, as conventional practices often lead to N losses, greenhouse gas emissions, and reduced eco-efficiency. To address these challenges, the “Pusa N Doctor” app was developed using dark green colour index (DGCI) for precision N management in maize. The app was further validated in experiment conducted with three N rates- 0 kg/ha (N0PK), 50 kg/ha (N0PK), and 75 kg/ha (N75PK) as basal, along with two splits of N at 35 and 45 DAS as per app (N50PK+App and N75PK+App) and GSTM (N50PK + GSTM and N75PK+GSTM). The plant height, leaf area index, and plant N concentration was highest in N75PK+App. The highest crop growth rate between 0-30 DAS was observed in the N75PK+App treatment (9.97 g/m²/day). Conversely, the maximum relative growth rate between 30-60 DAS was in the N50PK+App, while the lowest was in N75PK+App. The highest harvest index of 35.13% was in N50PK+App. Except for N75PK+App and recommended dose of fertilizer (RDF), the partial N balance was close to 1, with a minimum value of 0.87 in N75PK+App. The lowest virtual N was in N50PK+App (0.45), while in N75PK+App it was 2.16 times higher than RDF. All N fertilized treatments except N50PK+App witnessed increased cost of cultivation over RDF. N50PK+App had 29.5% lower GHGI of N2O, with 11.6% and 13.3% higher energy and GHG-based eco-efficiency respectively than RDF. Thus, applying 50 kg N as basal along with its 2 splitting as per Pusa N Doctor, optimizes maize-growth, N use efficiency, eco-efficiency, and reduces GHG emissions.
Citation: Sarkar S, Upadhyay PK, Dey A, Ekka U, Rathore SS, Shekhawat K, et al. (2025) Integrating soil and crop metrics with precision agriculture: Pusa N Doctor app for sustainable nitrogen management in maize. PLoS ONE 20(4): e0318678. https://doi.org/10.1371/journal.pone.0318678
Editor: Abhay Omprakash Shirale, ICAR National Bureau of Soil Survey & Land Use Planning, INDIA
Received: December 26, 2024; Accepted: January 19, 2025; Published: April 2, 2025
Copyright: © 2025 Sarkar et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data relevant to this study are included in the paper.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: SMW, standard meteorological week; Tmax, maximum temperature; Tmin, minimum temperature; RH (E), Relative humidity evening; RH (M), Relative humidity morning
1. Introduction
Despite notable advancements in agricultural practices, the challenge of imbalanced fertilizer application continues to persist in Indian agriculture, particularly due to a pronounced reliance on nitrogen (N) fertilizers [1]. Farmers, often driven by the perceived benefits of increased yields, tend to apply nitrogen disproportionately, neglecting the balanced use of other essential nutrients. This skewed fertilization practice not only reduces nutrient use efficiency but also imposes a substantial financial burden on farmers, as excessive nitrogen application often results in diminishing economic returns [2,3]. Moreover, the environmental repercussions of such practices are profound. Over-application of nitrogen contributes to nitrate leaching into groundwater, posing risks to water quality and public health. Elevated nitrate levels in drinking water are linked to severe health issues, including methemoglobinemia in infants [4,5]. Additionally, excessive nitrogen disrupts soil health by altering its chemical and biological properties, leading to long-term productivity decline [6]. These environmental consequences, coupled with reduced nutrient use efficiency, amplify the challenges faced by Indian agriculture. In the broader context, the cumulative impact of these issues undermines agricultural productivity, threatening the sustainability of food production systems and posing a significant risk to global food security [7].
Therefore, precision nutrient management (PNM), adhering to the 4R nutrient stewardship principles—applying the right fertilizer source, at the right rate, at the right time, and in the right place [8], is crucial for optimizing N use, directly impacting both economic returns and environmental sustainability. This approach aims to enhance crop productivity, boost farmer profitability, protect the environment, and improve both sustainability and nutrient use efficiency [9,10].
Several tools for real-time N management have been developed, such as GreenSeeker, SPAD (The Soil Plant Analysis Development) chlorophyll meter, and LCC (leaf colour chart), which predict crop N needs based on leaf reflectance and greenness [11,12]. These tools evaluate leaf N levels, which correlate with photosynthesis and biomass production, making them sensitive indicators of crop N demand throughout the growing season [13]. The GreenSeeker™ (GSTM) (Trimble Ltd., Sunnyvale, CA, USA) is particularly effective for making mid-season N recommendations based on the normalized difference vegetation index (NDVI). But the high cost of GSTM and SPAD 502Plus meter (Konica Minolta Inc., New Jersey, USA) and the need for complex calculations limit their adoption among Indian farmers [14, 15]. In contrast, the LCC is an affordable, user-friendly tool for monitoring leaf greenness as an indicator of N status [16,17]. However, LCC’s limitations include its binary N recommendations—whether to apply or not, without specifying the dosage—and its reliance on the farmer’s subjective colour perception [18]. According to a 2018 study by US media agency Zenith, with over half of India’s population owning camera-equipped smartphones, mobile-based applications offer a promising digital platform to estimate leaf N status. Therefore, an attempt has been made to develop a smartphone app “Pusa N Doctor”, based on the dark green colour index (DGCI) [19], which can provide real-time N recommendation in maize by analyzing images of crop leaves taken with a smartphone camera. This study aims to validate the app, with respect to maize growth parameters, yield attributes, financial advantage, N use efficiency, nitrous oxide (N2O) emission, and eco-efficiency.
2. Materials and methods
2.1. Study site and climatic conditions
Field experiments were conducted with maize during 2020, 2021 and 2022 in the experimental field of ICAR- Indian Agricultural Research Institute (ICAR-IARI), New Delhi (28˚4′ N, 77˚12′ E, 228.6 m above sea level). The study area experiences a subtropical, semi-arid climate, marked by hot, dry summer and cold winter. More details regarding the weather parameter during study period is presented in Fig 1. The soil at the experimental site was sandy loam (Typic Haplustept) with a slightly alkaline pH of 7.7. The initial soil properties comprised of a mildly alkaline soil reaction (1:2.5 soil-to-water) of 8.1, an EC of 0.35 dS/m, low organic carbon content of 0.44%, low available N at 185 kg/hectare, and medium levels of available phosphorus (P) and potassium (K), measuring 12.8 kg/hectare and 205 kg/hectare, respectively. Before the initiation of the validation experiment in the kharif 2023, initially the top 0-15 cm soil contained 0.49% organic carbon, 223 kg/ha available N, 16.8 kg/ha available P, and 242 kg/ha available K.
2.2. App development
2.2.1. Treatment details.
The three-season experiment for the app development during 2020, 2021, and 2022 were conducted with six popular varieties/hybrids of maize crop (PC3, PC4, AH-4271, DKC-9164, PJMH-1, PMH-1) along with seven levels of N application, i.e., 0, 40, 80, 120, 160, 200 and 240 kg/ha in a factorial randomized complete block design (RCBD) replicated five times. A basal dose of 75 kg/ha of P2O5 and K2O was added in all the plots irrespective of the treatments. In the treatment with 40 kg/ha N, the entire N amount was applied as a basal dose. For the 80 kg/ha N treatment, 40 kg/ha N was applied as a basal dose and an additional 40 kg/ha N at 35 days after sowing (DAS). For 120 kg/ha N and above, one-third of N was used as basal, one-third at 35 DAS, and one-third at 45 DAS.
2.2.2. Smartphone app development.
The app development process is depicted in Fig 2. The DGCI was calculated [19] as:
Here, DGCI is the dark green colour index; RGB is red, green and blue colour space; HSB is the hue, saturation and brightness colour space; and INSEY is the in-season estimated yield.
S = C/ max (R, G, B)
H = 60×(G-B)/C (when, max (R, G, B) = R)
H = 60×(2+(B-R)/C) (when, max (R, G, B) = G)
H = 60×(4+(R-G)/C) (when, max (R, G, B) = B)
C = max (R, G, B)-min (R, G, B)
B = max (R, G, B)
In the above equation, H: hue, S: saturation, B: brightness; C: chroma; and R, G, B denotes the normalized red, green, blue levels respectively.
The INSEY which is the in-season estimated yield was computed as:
Where, GDD is the growing degree days.
Response Index (RI) was deduced according to the below equation:
Where, DGCI (ENRICH) is the DGCI value of the N enriched strip maintained by supplying 240 kg N/hectare.
The empirically derived power function relating INSEY and grain yield at 35 and 45 DAS is represented as:
Where, YP0 is the potential yield without additional fertilizer; a, and b are constants.
The achievable yield with added N fertilizer to the test plots (YPN) was assessed by multiplyingYP0 by the RI as described below:
Finally, fertilizer N prescription was calculated as follows:
Where, 1.43 represents the mean maize grain N concentration and 0.5 is the achievable efficiency factor in South-Asia [20].
In the validation experiment, Pusa N Doctor validated against GSTM, and for the calculation of N dose with GSTM, the DGCI values were simply replaced by the NDVI values in equations (2) and (3).
2.3. Validation experiment
2.3.1. Treatment details and crop management.
The validation experiment was conducted during kharif 2023 in RCBD with 6 treatments and 4 replications. The treatments comprised of 0 kg/ha (N0PK), 50 kg/ha (N50PK) and 75 kg/ha (N75PK) of N, in addition to 75 kg of P2O5 and K2O incorporated as basal (Table 1). Additionally, N doses of 41.9 kg/ha and 30 kg/ha were applied in two splits at 35 and 45 DAS respectively following precision N management-based recommendations through android-based Pusa N Doctor application in N50PK+App while 58.6 kg/ha and 41.2 kg/ha were applied in N75PK+App. In the GreenSeeker™ based treatments, two splits of 39.4 kg/ha and 36.4 kg/ha of N were applied respectively at 35 and 45 DAS in N50PK+GSTM, whereas 35.6 kg/ha and 33.7 kg/ha were applied in N75PK+GSTM. These treatments were compared with the recommended dose of fertilizer (RDF), where 150 kg/ha of N was used in equal splits as basal, at 30 DAS and 50 DAS.
Initial soil properties of the validation experiment are given in Table 2. In all experimental treatments, PJMH1 maize cultivar, released jointly by ICAR-IARI, New Delhi and Jawaharlal Nehru Krishi Vishwa Vidyalaya (JNKVV), Jabalpur, was selected for cultivation. The land preparation process involved deep summer ploughing before monsoon arrival, followed by, one ploughing, one discing, one planking to ensure good tilth. To prevent surface water flow from one plot to another, bunds were built around each plot. The crop was sown using tractor drawn opener with 60 cm x 20 cm spacing, and 20 kg/ha seed rate, on 4th July, 2023. During the entire cropping season three irrigations of 5 cm depth each were given on 3rd August (30 DAS), 18th August (47 DAS), and 30th August (57 DAS), 2023, when there was a dry spell.
2.3.2. Measured and calculated parameters.
2.3.2.1.Growth parameters: At 30, 60, and 90 days after sowing (DAS) and at harvest, three plants were randomly selected from each plot to measure plant height. The measurements were taken using a meter scale, and the average height was recorded in centimetres (cm). During the tasselling and harvest stages, height was measured from the ground to highest fully open leaf, while at the knee-height stage, the measurement extended from the ground to the tip of the folded leaf.
Leaf area was measured at 30 and 60 days after sowing (DAS) by removing the leaves from the same three maize plants previously selected for dry matter analysis. The leaves were then measured using a leaf area meter (Model: LI-COR-3100). This data was used to calculate the leaf area index (LAI) using the following formula:
Dry matter accumulation was monitored at 30 DAS, 60 DAS, 90 DAS and at harvest by pulling out three maize plants per plot. The plants were initially air-dried for 7 to 8 days, followed by oven drying at 65°C for two days until a stable mass was achieved, which was recorded as grams of dry matter/ plant. The crop growth rate (CGR) and relative growth rate (RGR) was then worked out by the following formula:
The relative growth rate was calculated out with the following formula [21]:
Where,
ln: the natural logarithm, W1 and W2 are dry weight (g) of plants at time T1 and T2, respectively.
T2 – T1 is the interval of time in days.
A is the land area (m2) occupied by plants
2.3.2.2.Shelling percentage and harvest index: To determine crop yield, the harvested area excluded two border rows on each side and 0.5 m strip along the length of the plot. Once the stover was separated and the husk along with the silk were removed, cobs from every net plot were sun-dried and the weight was expressed as cob yield (t/ha). The maize stover was cut at ground level and also sun-dried before being weighed. The total weight of the harvested materials (both cobs and stover) from each net plot was measured and reported as biological yield (t/ha). Cobs from each net plot were shelled, and the grains were sun-dried. The final grain yield was then adjusted to 14.5% moisture content and expressed in tons/hectare (t/ha).
Shelling percentage was calculated as follows:
The harvest index was calculated using the following equation:
2.3.2.3.Plant analysis: After sun-drying, plant samples collected at 40 and 50 DAS, harvested maize grain and stover samples were further dried in a hot air oven at 60°C for 24 hours to reach a constant weight. The dried samples were then ground using a Retsch mixer mill MM 400 and passed through a 40-mesh sieve. Further, 0.5 g of sample from the plots were used to determine N concentration. N content (%) in maize plant at 40 and 50 DAS, grain and stover was analyzed using the modified Kjeldahl method [22] by a CHNS analyser (Carbon, Hydrogen, Nitrogen, and Sulfur analyser) (Euro EA-3000). The N uptake by maize grain (Grain N uptake), and stover (Stover N uptake) was calculated by multiplying the N content (%) in maize grain with grain yield (kg/ha). Total N uptake was finally expressed as the sum of grain N uptake and stover N uptake.
2.3.2.4.Nitrogen use efficiency: Following formulae were used for the computation of physiological efficiency of N (PEN) [23]:
The partial N balance (PNBN) was computed as per the following equation (24):
The internal utilization efficiency of N (IUEN) was calculated as per Parco et al. [25]; Qiu et al. [26]:
The virtual N factor (VNF) was calculated by using following formula [27]:
Where the N loss to the environment was calculated as the difference between N input (fertilizer + soil N supply) and N output (Total plant N uptake). The total N uptake by the no N control plot was assumed as the soil N supply.
In Equations (13) and (14), Total N uptakeN represents the total N uptake by the maize stover and grain from the plots applied with N (kg/ ha), Total N uptakeC represents the total N uptake by the maize stover and grain from the control (kg/ ha), FN denotes the dose of N used (kg/ ha), Grain yieldN denotes grain yield from the plots applied with N (kg/ha), Grain yieldC denotes grain yield from control plot and Grain N UptakeN represents the grain N uptake from the plots applied with N (kg/ ha).
2.3.2.5.Economic analysis: The calculation of the cost of cultivation (CoC) was done based on current input prices. The total cost accounted for various inputs like fertilizers, irrigation, seeds, and agronomic activities, including soil cultivation, sowing, pest control, and harvesting. Human labour costs were computed based on an eight-hour workday, as outlined by the Indian labour laws in addition to hours/hectare of machinery time for each farm activity. Total costs were determined by summing labour, machinery time, fuel, and electricity for each activity. The rental value of the land was also factored into the cultivation costs. The cost of GSTM was considered as US $ 960, with life span of 16 years, which amounts to additional US $ 60 for a single year for the GSTM based treatments. Saving of cost over RDF ($/ha) was computed as the difference in CoC between RDF and the rest of the treatments. The economic assessments were carried out in Indian rupees and subsequently transformed to US $.
2.3.2.6.Greenhouse gas intensity of nitrous oxide and eco-efficiency index: Greenhouse gas (GHG) emitted by the treatments were assessed using the Cool Farm Tool (CFT v1.7.1) [28], incorporating multiple globally validated models into a unified system to estimate emissions resulting from agricultural inputs in crop production systems. The Greenhouse gas intensity (GHGI) of N2O was calculated as:
Eco-efficiency integrates both the financial and environmental dimensions of a production system. It measures the effectiveness of a cropping system in boosting economic output while minimizing environmental impacts, such as energy consumption (equation (17)) and GHG emissions (equation (18)) from agriculture [29, 30].
Energy input was categorized into direct (labour, fuel, electricity) and indirect (seeds, fertilizers, machinery, chemicals) energy inputs. It was further divided into renewable (seeds, labour) and non-renewable sources (fuel, electricity, machinery, fertilizers, chemicals). Energy consumption was determined by applying energy coefficients [31, 32, 33, 34] to inputs.
2.4. Statistical analysis
The data on various parameters were analyzed using analysis of variance (ANOVA) for RCBD. The differences between treatment means were tested at P < 0.05 according to Fisher’s LSD test [35] and to depict them boxplots were generated for harvest index, N use efficiency indices, and GHGI. Ridgeline plot was developed for the eco- efficiency index, where the black line within the plot represents the mean, the different lowercase letters within the plot correspond to the significant difference of the treatments at p < 0.05 according to LSD test, purplish colour indicates lower eco- efficiency index value, and as we move towards yellowish tinge, we can find a comparatively higher eco- efficiency index. All the statistical analysis was done in R programming (version 4.2.1) and MS-Excel (Microsoft corporation, 2019).
3. Results
3.2. Plant height and leaf area index
There was significant effect of N management practices on the plant height of maize at 60 DAS, 90 DAS and at harvest (Table 3). The maize plant height in different treatments were at par with each other at 30 DAS. The highest maize height was in N75PK+App treatment at 30, 60, 90 DAS and at harvest. At 60 DAS, though plant height was highest under N75PK+App but it has no significant difference with RDF, followed by N75PK+GSTM, N50PK+GSTM. At 90 DAS and harvest, the treatments had a comparable height, which was significantly higher than the control.
The LAI of maize was significantly influenced by the N management practices with the highest being under N75PK+App treatment at both 30 DAS (1.40) and 60 DAS (4.20) (Table 4). There was no significant difference in LAI under N75PK+App and N75PK+GSTM, but it was 18.6% higher than RDF at 30 DAS. But, at 60 DAS, N75PK+App and N75PK+GSTM had a comparable LAI with RDF. Though the N50PK+App had a comparatively lower LAI, but it was statistically at par with N50PK+GSTM.
3.3. Crop growth rate, relative growth rate, shelling percentage and harvest index
There was a significant difference amongst the N fertilized plots with respect to the CGR and RGR of maize (Table 5). Though between 0-30 days’ time interval the highest CGR (9.97 g/m2/day) was in N75PK+App treatment, it was at par with N75PK+GSTM (8.86 g/m2/day) and RDF (9.45 g/m2/day), which again had no significant difference with N50PK+App (8.50 g/m2/day) and N50PK+GSTM (8.67 g/m2/day). On the contrary, the maize RGR between 30-60 DAS in N50PK+App was the maximum, while N75PK+App had the lowest RGR of 0.0351 g/g/day. Between 30-60 DAS and 60-90 DAS there was at par CGR in the N fertilized plots which differed significantly with the control. Similar was the result for the RGR at 60-90 DAS as well.
The shelling percent (Fig 3) in maize was highest in N50PK+App (76.88%), which was statistically at par with the control (67.75%) and other N fertilized plots. The lowest harvest index (Fig 4) was in the control plot (27.6%), followed by N75PK+App treatment (33.3%), which had no statistically significant difference with the other precision N based treatments including N50PK+App (35.13%) which had the highest harvest index.
3.4. Nitrogen content in plant at 40 and 50 DAS, and in grain and stover at harvest
There was no statistically significant difference between the treatments with respect to their plant N (%) at both 40 and 50 DAS (Table 6). The highest N content in maize plant at 40 DAS was in the plots applied with 75 kg N as basal (N75) N75PK + App (2.79%) and N75PK + GSTM (2.71%). Similar was the trend for 50 DAS as well, where, N75 had the maximum N % of 1.69% followed by RDF (1.67%) and the plots applied with 50 kg N as basal (N50) (1.65%).
All the Precision N based treatments were at par with the RDF in their grain and stover N content (Table 6). N75PK + App had the highest grain and stover N of 1.743% and 0.479% respectively, followed by N75PK + GSTM (grain and stover N content of 1.736% and 0.473% respectively). The grain and stover N content progressively decreased for RDF, N50PK + GSTM, and N50PK + App.
3.5. Nitrogen use efficiency
The N management practices did not influence the PEN significantly. The highest physiological efficiency was in N50PK+App (44.5), but was at par with N75PK+App (40.15) and other N fertilized treatments (Fig 5). Except N75PK+App and RDF, all the N fertilized plots had PNB greater than but close to 1 (Fig 6). The PNB in N50PK+App (1.19) had no significant difference with N50PK+GSTM (1.12), while the lowest PNB was in N75PK+App (0.87). All treatments had statistically comparable IUEN (Fig 7), though it was highest in N50PK+App. N75PK+App had the lowest IUEN. N50PK+App had the lowest VNF of 0.45 (Fig 8) which was comparable with RDF (0.38). This was followed by N50PK+GSTM and N75PK+GSTM having VNF of 0.63 and 0.67 respectively. The N75PK+App had the highest VNF, which was 2.16 and 1.8 times the value under RDF and N50PK+App respectively.
3.6. Saving of cost over RDF, greenhouse gas intensity of nitrous oxide and eco-efficiency index
Among the N fertilized plots, N50PK+App witnessed the maximum saving of 4 US $/ha over RDF. All other treatments except the control had increased CoC over RDF (Fig 9). The GHGI of N2O was lowest in N50PK+App (Fig 10), which was 29.5% lower than RDF. N50PK+App was followed by N50PK+GSTM (59.3 kg CO2-eq/t). N75PK+App had the highest GHGI of N2O, being significantly (15.9%) higher than RDF, followed by N75PK+GSTM (66.9 kg CO2-eq/t).
N50PK+App had the highest energy based eco-efficiency index of 0.116 US $/ MJ, which was 11.6% higher than RDF was statistically at par with RDF, N50PK+GSTM and N50PK+App (Fig 11). The control plot had the lowest energy based eco-efficiency index of 0.094 US $/ MJ, which had no significant difference with N75PK+App (0.098 US $/ MJ). Similar was the trend for GHG emission based eco-efficiency index as well, where the highest value was in N50PK+App (0.85 US $/ kg CO2 eq) which had no significant difference with N50PK+GSTM (0.78 US $/ kg CO2 eq) (Fig 12). This was followed by N75PK+GSTM (0.76 US $/ kg CO2 eq) and RDF (0.75 US $/ kg CO2 eq), progressively decreasing in N75PK+App (0.72 US $/ kg CO2 eq), with the lowest value in control (0.55 US $/ kg CO2 eq).
4. Discussion
4.1. Crop growth parameters
The rate of N application is highly correlated with the growth parameters of a crop [36, 37, 38, 39]. In our study, the highest cumulative dose of N was applied in N75PK+App treatment, which improved the maize plant height, LAI as well the dry matter accumulation. Similar to our findings, Hammad et al. [40], reported a higher maize plant height and dry matter accumulation with a higher dose of N application in maize. The addition of N boosted cell counts and leaf volume, sped up chlorophyll production, and enhanced plant biomass in the early stages of crop development [41]. Moreover, Amanullah et al. [42], also observed a positive correlation between LAI and fertilizer N dose in maize. Similar results have been reported by Amanullah and Shah [43], Ning et al. [44], Zhang et al. [45], Patra et al. [39], Deng et al. [46], who also found enhanced growth, LAI, photosynthetic rate and biomass accumulation in maize with an increment in the N fertilizer application rate.
While the CGR at (0-30) DAS was maximum in N75PK+App, but the RGR followed an opposite trend where the N50PK+App had the highest RGR and it was least in N75PK+App. This might be because, a small plant that is growing relatively quickly can exhibit high RGR at its early stages, as RGR depends on the initial biomass of the plant and the relative increase in biomass is significant [47, 48]. On the other hand, CGR is more influenced by total biomass increase per unit area and time, which might be low resulting in a lower CGR [49, 50].
4.2. Shelling percentage and harvest index
The highest shelling percent in maize was achieved in N50PK+App, which was statistically at par with the other N fertilized plots and the no N control. The shelling percentage in maize may not improve with N application compared to control plot because shelling percentage is primarily influenced by the ratio of grain to cob. N application often boosts overall plant growth, including both grain and cob development, but the relative increase in grain weight might be proportionate to the cob, leading to little or no change in the shelling percentage. The lowest harvest index was in the control plot, followed by N75PK+App treatment, which had no statistically significant difference with the other precision N based treatments and RDF. Similar to our finding, Zhang et al. [45], found at par harvest index across the different N application treatments. Moreover, field experiment on maize in Ethiopia showed that increasing N rates consistently boosted the harvest index until reaching the maximum application of 115 kg N/ha [51].
4.3. Nitrogen content in plant at 40 and 50 DAS, and in grain and stover at harvest
At Both 40 and 50 DAS the N content in maize plant as well as the N content in maize grain and stover at harvest was highest in the treatments applied with 75 kg N as basal, while the lowest values were in the treatments applied with 50 kg N as basal, though all the treatments were statistically at par. There was no significant difference amongst the N fertilized plots with that of the no N control due to a dilution effect: as N fertilizer rates increase, so does dry matter production, which dilutes the N concentration in the plant [52, 53]. With lower dry matter production, the N content is relatively higher per unit of dry matter [54, 55]. Therefore, even though N application increases both dry matter and N uptake, the N concentration per 100 kg of dry matter remains similar between the control and the N-fertilized plots [56]. Mondal et al. [57] also noted that the N content in maize and stover was comparable, but the stover yield varied significantly across different fertilizer N treatments. The varying N application rates of 0, 120 and 150 kg N/ ha did not lead to statistically significant differences in the N content of the grain between treatments in a subsurface drip fertigated conservation agriculture system [39].
4.4. Nitrogen use efficiency
In our study, all the treatments were at par with each other in terms of their physiological efficiency of N. Physiological efficiency is the efficiency with which absorbed N is converted into biomass or grain [58]. This metric is determined by the plant’s inherent capacity to use the absorbed N for growth and development [24]. Since physiological efficiency is a function of the plant’s internal metabolism and indicates the proportional increase in yield with increase in N plant uptake, it remained stable across different fertilizer treatments [59, 60].
Typically, a PNB value greater than 1 indicates that more nutrients are being extracted from the soil through crop harvesting than are being replenished by fertilizers or manure [61, 62]. This imbalance can result in the depletion of soil nutrients, a process known as “soil mining.” In our study, PNB was greater than but close to 1 in most of the N fertilized plots except N75PK+App and RDF. This can be because PNB is a partial balance and does not consider the indigenous soil N supply [63], thereby indicating that a PNB greater than 1 might not always lead to soil N mining. The lowest PNB was in N75PK+App, which is lower than 1, indicates a significant amount of N loss, due to application of excess amount of N than the crop requirement, further leading to environmental deterioration [64].
The IUEN is relatively stable across different N fertilized treatments. Since IUEN is the ratio of grain yield to total N uptake, it implies that both the grain yield and N uptake increase or decrease proportionally [25]. As a result, even though there may be variations in grain yield and N uptake between treatments, the efficiency itself remains consistent, which is why no statistically significant difference is observed across treatments. Similar to our result, Qiu et al. [26], in field experiments in three villages of Jilin, China, observed that IUEN showed little variation when N was applied at rates of 140, 210, and 280 kg/ha. The highest IUEN in N50PK+App, with least value in N75PK+App, is due to the moderately high grain yield with a low total N uptake [65] in N50PK+App, and highest grain yield and total N uptake in N75PK+App.
The N75PK+App had the highest VNF, which was mainly attributed to its highest N loss, significantly higher than other N fertilized plots. The lowest VNF in RDF, which was at par with N50PK+App, can be explained by a high grain N uptake, with low N loss.
4.5. Saving of cost over RDF, greenhouse gas intensity of nitrous oxide and eco-efficiency index
Saving of cost over RDF was highest in control plot followed by N50PK+App, and all other treatments had increased cost over RDF. This was because of no nitrogenous fertilizer application and the lowest dose of fertilizer N application in control and N50PK+App respectively. The remaining treatments had higher CoC compared to RDF, because of the highest dose of N fertilizer application in N75PK+App, and consideration of the cost of the GSTM during CoC computation in the GSTM based treatments.
The GHGI of N₂O was highest in the N75PK+App treatment, while the lowest value was recorded in the N50PK+App treatment. Applying a higher dose of N enhances the soil N content. This surplus N is more likely to undergo various microbial activities, like nitrification and denitrification. During these processes, microbes convert excess N into N₂O, a potent GHG [66]. Halli et al. [34], delineated that the use of nitrogenous fertilizer had a direct response to N₂O emission rate. Specifically, applying N in both basal and split doses at 100 and 120 kg/ha, respectively, led to higher N₂O emissions of 2.14 kg/ha, compared to 1.08 kg/ha in untreated plots. In the maize-wheat cropping system, N fertilizer was the chief causal factor to direct N₂O emissions, making up 65.9% of the cumulative emissions [67]. Chai et al. [68] determined that the N fertilizer use in wheat and maize resulted in 35.82 Gg/year and 69.44 Gg/year N₂O emission respectively in China. A study conducted over two years at five farms in Michigan, USA, investigated maize with six rates of N ranging between 0 to 225 kg/ha per season. The research consistently found that N₂O emissions increased exponentially with higher N rates at each site, regardless of the year [69, 70]. Moreover, a global meta-data analysis, comprising of 78 studies spanning 233 sites, found that the N₂O emissions rose exponentially once the N rate surpassed the crop requirement [71, 72].
N50 based treatments (N50PK+App and N50PK+GSTM) showed the highest energy based as well as GHG emission based eco-efficiency index, with lowest value in N75PK+App. This trend was opposite to the N dose applied to the crop. In the overall energy expenditure in agricultural systems, fertilizer, being the most energy-demanding input [73, 74], contributed the largest share due to its non-renewable nature [32], and accordingly led to the lowest energy based eco-efficiency index value in N75PK+App, applied with the highest N dose. Analogous to this result, the increased application of N fertilizers also has a higher carbon footprint due to the energy-intensive character of fertilizer manufacture, transport, application [33], along with nitrous oxide emission from soil, which attributed to the lowest GHG emission based eco-efficiency index in N75PK+App. A study in Southern Italy found that higher yields and economic returns don’t always offset the increased resource use, reducing eco-efficiency [75]. Low-resource farming, especially with minimal N use, proved more eco-efficient than intensive methods [76].
5. Conclusion
The Pusa N Doctor app, utilizing a DGCI-based algorithm, effectively optimized nitrogen management in maize by improving N use efficiency, reducing GHG emissions, and enhancing eco-efficiency. The app demonstrated superior results in comparison to traditional tools like GreenSeeker™ particularly in terms of cost-effectiveness and precision in N management. By recommending a basal N application of 50 kg with split applications, it ensures sustained maize growth, efficient N use, and a reduction in environmental impacts, making it a promising alternative to existing nitrogen management tools. Furthermore, the implementation of such innovative digital tools has the potential to revolutionize precision agriculture by providing scalable and farmer-friendly solutions. Future research could focus on refining the app for diverse cropping systems and agro-climatic conditions, while policy integration could enhance adoption rates, ultimately contributing to sustainable agriculture and climate change mitigation on a global scale.
Acknowledgments
The authors sincerely thank the Director of IARI for providing the necessary facilities for this investigation.
References
- 1. Velayudhan PK, Sivalingam N, Jha GK, Singh A, Pathak H. Nitrogen budget of Indian agriculture: trends, determinants and challenges. Environ Dev Sustain. 2023;26(4):10225–42.
- 2. Upadhyay PK, Dey A, Singh VK, Dwivedi BS, Singh T, G A R, et al. Conjoint application of nano-urea with conventional fertilizers: An energy efficient and environmentally robust approach for sustainable crop production. PLoS One. 2023;18(7):e0284009. pmid:37406009
- 3.
Fertiliser Statistics (2022–23) . Fertiliser Association of India, New Delhi. 2022.
- 4. Aulakh MS. Integrated nutrient management for sustainable crop production, improving crop quality and soil health, and minimizing environmental pollution. 19th World Congress of Soil Science, Soil Solutions for a Changing World. 2010; 1–6. https://www.researchgate.net/publication/320044606
- 5. Zebarth BJ, Ryan MC, Graham G, Forge TA, Neilsen D. Groundwater monitoring to support development of BMPs for groundwater protection: The Abbotsford‐Sumas Aquifer Case Study. Groundwater Monitoring Rem. 2015;35(1):82–96.
- 6.
Tayoh LN. Destruction of Soil Health and Risk of Food Contamination by Application of Chemical Fertilizer. In: Bauddh K, Kumar S,Singh RP,Korstad J, editors. Ecological and Practical Applications for Sustainable Agriculture. Singapore: Springer Singapore; 2020. p. 53–64. https://doi.org/10.1007/978-981-15-3372-3_3
- 7. Pingali PL. Green revolution: impacts, limits, and the path ahead. Proc Natl Acad Sci U S A. 2012;109(31):12302–8. pmid:22826253
- 8.
Force IT. The global “4R” nutrient stewardship framework. Developing fertilizer best management practices for delivering economic, social, and environmental benefits. IFA Task Force on Fertilizer Best Management Practices. International Fertilizer Industry Association (IFA): Paris, France. 2009.
- 9. Dwivedi BS, Singh VK, Meena MC, Dey A, Datta SP. Integrated nutrient management for enhancing nitrogen use efficiency. Indian J Fertil. 2016;12:62–71.
- 10. Bruulsema TW, Peterson HM, Prochnow LI. The Science of 4R Nutrient Stewardship for Phosphorus Management across Latitudes. J Environ Qual. 2019;48(5):1295–9. pmid:31589734
- 11. Bijay-Singh, Varinderpal-Singh, Yadvinder-Singh, Kumar A, Jagmohan-Singh, Vashistha M, et al. Fertilizer Nitrogen management in irrigated transplanted rice using dynamic threshold greenness of leaves. Agric Res. 2016;5(2):174–81.
- 12. Mitra B, Singha P, Roy Chowdhury A, Sinha AK, Skalicky M, Laing AM, et al. Optical sensor-based nitrogen management: an environmentally friendly and cost-effective approach for sustainable wheat (Triticum aestivum L.) production on Eastern plains of India. Front Sustain Food Syst. 2023;7:1153575.
- 13. Varinderpal-Singh K, Gosal SK, Choudhary R, Singh R, Adholeya A. Improving nitrogen use efficiency using precision nitrogen management in wheat (Triticum aestivum L.). J Plant Nutr Soil Sci. 2021;184(3):371–7.
- 14. Crain J, Ortiz-Monasterio I, Raun B. Evaluation of a Reduced Cost Active NDVI Sensor for Crop Nutrient Management. Journal of Sensors. 2012;2012:1–10.
- 15. Rehman TH, Borja Reis AF, Akbar N, Linquist BA. Use of Normalized Difference Vegetation Index to Assess N Status and Predict Grain Yield in Rice. Agronomy Journal. 2019;111(6):2889–98.
- 16. Singh V, Singh B, Thind HS, Singh Y, Gupta RK, Singh S. Evaluation of leaf colour chart for need-based nitrogen management in rice, maize and wheat in north-western India. J Res Punjab Agric Univ. 2014;51(3 & 4):239–45.
- 17. Tao M, Ma X, Huang X, Liu C, Deng R, Liang K, et al. Smartphone-based detection of leaf color levels in rice plants. Computers and Electronics in Agriculture. 2020;173:105431.
- 18. Yang W-H, Peng S, Huang J, Sanico AL, Buresh RJ, Witt C. Using leaf color charts to estimate leaf nitrogen status of rice. Agronomy Journal. 2003;95(1):212–7.
- 19. Karcher DE, Richardson MD. Quantifying Turfgrass Color Using Digital Image Analysis. Crop Science. 2003;43(3):943–51.
- 20.
Singh Y, Singh B, Ladha JK, Singh JP, Choudhary OP. Enhancing nitrogen use efficiency for sustainable rice–wheat production system in the Indo-Gangetic Plains of India. Agricultural Nitrogen Use and its Environmental Implications. New Delhi: IK International Publishing House; 2007, p. 139–164
- 21. Blackman VH. The Compound Interest Law and Plant Growth. Annals of Botany. 1919;os-33(3):353–60.
- 22.
Jackson ML. Soil chemical analysis. New Delhi, India: Prentice Hall of India Pvt, 1973, pp. 151–154.
- 23. Furoc RE, Morris RA. Apparent recovery and physiological efficiency of nitrogen in sesbania incorporated before rice. Agronomy Journal. 1989;81(5):797–802.
- 24.
Dobermann A. Nutrient use efficiency–measurement and management. Fertilizer Best Management Practices. Paris: International Fertilizer Industry Association (IFA). 2007:1–19.
- 25. Parco M, Ciampitti IA, D’Andrea KE, Maddonni GÁ. Prolificacy and nitrogen internal efficiency in maize crops. Field Crops Research. 2020;256:107912.
- 26. Qiu SJ, He P, Zhao SC, Li WJ, Xie JG, Hou YP, et al. Impact of nitrogen rate on maize yield and nitrogen use efficiencies in northeast China. Agronomy Journal. 2015;107(1):305–13.
- 27. Galloway JN, Winiwarter W, Leip A, Leach AM, Bleeker A, Erisman JW. Nitrogen footprints: past, present and future. Environ Res Lett. 2014;9(11):115003.
- 28. Hillier J, Walter C, Malin D, Garcia-Suarez T, Mila-i-Canals L, Smith P. A farm-focused calculator for emissions from crop and livestock production. Environmental Modelling & Software. 2011;26(9):1070–8.
- 29. Cicek A, Altintas G, Erdal G. Energy consumption patterns and economic analysis of irrigated wheat and rainfed wheat production: case study for Tokat region, Turkey. Bulg J Agric Sci. 2011;17(3):378–88.
- 30. Gómez-Limón JA, Picazo-Tadeo AJ, Reig-Martínez E. Eco-efficiency assessment of olive farms in Andalusia. Land Use Policy. 2012;29(2):395–406.
- 31. Babu S, Singh R, Avasthe RK, Yadav GS, Rajkhowa DJ. Intensification of maize (Zea mays) –based cropping sequence in rainfed ecosystem of Sikkim Himalayas for improving system productivity, profitability, employment generation and energy-use efficiency under organic management condition. Indian J Agri Sci. 2016;86(6):86.
- 32. Jat SL, Parihar CM, Singh AK, Kumar B, Choudhary M, Nayak HS, et al. Energy auditing and carbon footprint under long-term conservation agriculture-based intensive maize systems with diverse inorganic nitrogen management options. Sci Total Environ. 2019;664:659–68. pmid:30763846
- 33. Yadav GS, Das A, Kandpal BK, Babu S, Lal R, Datta M, et al. The food-energy-water-carbon nexus in a maize-maize-mustard cropping sequence of the Indian Himalayas: an impact of tillage-cum-live mulching. Renewable and Sustainable Energy Reviews. 2021;151:111602.
- 34. Halli HM, Govindasamy P, Wasnik VK, Shivakumar BG, Swami S, Choudhary M, et al. Climate-smart deficit irrigation and nutrient management strategies to conserve energy, greenhouse gas emissions, and the profitability of fodder maize seed production. Journal of Cleaner Production. 2024;442:140950.
- 35.
Gomez KA, Gomez AA. Statistical procedures for agricultural research. 2nd ed. New York: John Wiley and Sons; 1984, p. 680.
- 36. , Iqbal A, Ali A, Fahad S, Parmar B. Nitrogen source and rate management improve maize productivity of smallholders under semiarid climates. Front Plant Sci. 2016;7:1773. pmid:27965685
- 37. Abdou NM, Abdel-Razek MA, Abd El-Mageed SA, Semida WM, Leilah AAA, Abd El-Mageed TAA, et al. High nitrogen fertilization modulates morpho-physiological responses, yield, and water productivity of lowland rice under deficit irrigation. Agronomy. 2021;11(7):1291.
- 38. Galdi LV, Cordeiro CF dos S, de Senna e Silva B, de La Torre EJR, Echer FR. Interactive effects of increased plant density, cultivars and N rates in environments with different cotton yield recovery potential. Industrial Crops and Products. 2022;176:114394.
- 39. Patra K, Parihar CM, Nayak HS, Rana B, Sena DR, Anand A, et al. Appraisal of complementarity of subsurface drip fertigation and conservation agriculture for physiological performance and water economy of maize. Agricultural Water Management. 2023;283:108308.
- 40. Hammad HM, Chawla MS, Jawad R, Alhuqail A, Bakhat HF, Farhad W, et al. Evaluating the Impact of Nitrogen Application on Growth and Productivity of Maize Under Control Conditions. Front Plant Sci. 2022;13:885479. pmid:35685007
- 41. Elsayed S, El-Hendawy S, Elsherbiny O, Okasha AM, Elmetwalli AH, Elwakeel AE, et al. Estimating Chlorophyll Content, Production, and Quality of Sugar Beet under Various Nitrogen Levels Using Machine Learning Models and Novel Spectral Indices. Agronomy. 2023;13(11):2743.
- 42. , Amanullah , Almas LK, Shah P. Timing and rate of nitrogen application influence profitability of maize planted at low and high densities in northwest Pakistan. Agronomy Journal. 2010;102(2):575–9.
- 43. Amanullah, Shah P. Nitrogen rates and its time of application influence dry matter partitioning and grain yield in maize planted at low and high densities. Journal of Plant Nutrition. 2010;34:224–242.
- 44. Ning P, Fritschi FB, Li C. Temporal dynamics of post-silking nitrogen fluxes and their effects on grain yield in maize under low to high nitrogen inputs. Field Crops Research. 2017;204:249–59.
- 45. Zhang L, Liang Z, He X, Meng Q, Hu Y, Schmidhalter U, et al. Improving grain yield and protein concentration of maize (Zea mays L.) simultaneously by appropriate hybrid selection and nitrogen management. Field Crops Research. 2020;249:107754.
- 46. Deng T, Wang J-H, Gao Z, Shen S, Liang X-G, Zhao X, et al. Late Split-Application with Reduced Nitrogen Fertilizer Increases Yield by Mediating Source-Sink Relations during the Grain Filling Stage in Summer Maize. Plants (Basel). 2023;12(3):625. pmid:36771709
- 47.
Hunt R, Thomas B, Murphy DJ, Murray D. Growth analysis, individual plants. In: Encyclopedia of Applied Plant Sciences, 2nd ed. Elsevier: London. 2003:579–588.
- 48. Ben Youssef R, Boukari N, Abdelly C, Jelali N. Salicylic acid phytohormone as a new potential for adaptation and improvement of metabolic responses of fodder species under separate and combined impact of salinity and Fe deficiency constraints. Phytohormones and Stress Responsive Secondary Metabolites. 2023:249–63.
- 49. Alkahtani J, Elshikh MS, Alwahibi MS, Muhammad A, et al. Phosphorus and zinc fertilization influence crop growth rates and total biomass of coarse vs. fine types rice cultivars. Agronomy. 2020;10(9):1356.
- 50. Monzon JP, Cafaro La Menza N, Cerrudo A, Canepa M, Rattalino Edreira JI, Specht J, et al. Critical period for seed number determination in soybean as determined by crop growth rate, duration, and dry matter accumulation. Field Crops Research. 2021;261:108016.
- 51. Mosisa W, Dechassa N, Kibret K, Zeleke H, Bekeko Z. Effects of timing and nitrogen fertilizer application rates on maize yield components and yield in eastern Ethiopia. Agrosystems Geosci & Env. 2022;5(4).
- 52. Greenwood DJ, Lemaire G, Gosse G, Cruz P, Draycott A, Neeteson JJ. Decline in percentage N of C3 and C4 Crops with increasing plant mass. Annals of Botany. 1990;66(4):425–36.
- 53. Shao H, Miao Y, Fernández FG, Kitchen NR, Ransom CJ, Camberato JJ, et al. Evaluating critical nitrogen dilution curves for assessing maize nitrogen status across the US Midwest. Agronomy. 2023;13(7):1948.
- 54. Lemaire G, Jeuffroy M-H, Gastal F. Diagnosis tool for plant and crop N status in vegetative stage. European Journal of Agronomy. 2008;28(4):614–24.
- 55. Du L, Li Q, Li L, Wu Y, Zhou F, Liu B, et al. Construction of a critical nitrogen dilution curve for maize in Southwest China. Sci Rep. 2020;10(1):13084. pmid:32753694
- 56. Irmak S, Mohammed AT, Drudik M. Maize nitrogen uptake, grain nitrogen concentration and root-zone residual nitrate nitrogen response under center pivot, subsurface drip and surface (furrow) irrigation. Agricultural Water Management. 2023;287:108421.
- 57. Mondal S, Kumar R, Mishra JS, Dass A, Kumar S, Vijay KV, et al. Grain nitrogen content and productivity of rice and maize under variable doses of fertilizer nitrogen. Heliyon. 2023;9(6):e17321. pmid:37441387
- 58. Tabak M, Lepiarczyk A, Filipek-Mazur B, Lisowska A. Efficiency of nitrogen fertilization of winter wheat depending on sulfur fertilization. Agronomy. 2020;10(9):1304.
- 59. Galindo FS, Teixeira Filho MCM, Buzetti S, Rodrigues WL, Santini JMK, Alves CJ. Nitrogen fertilisation efficiency and wheat grain yield affected by nitrogen doses and sources associated withAzospirillum brasilense. Acta Agriculturae Scandinavica, Section B — Soil & Plant Science. 2019;69(7):606–17.
- 60. Mahboob W, Yang G, Irfan M. Crop nitrogen (N) utilization mechanism and strategies to improve N use efficiency. Acta Physiol Plant. 2023;45(4).
- 61. Brentrup F, Pallière C. Nitrogen use efficiency as an agro-environmental indicator. In: Proceedings of the OECD Workshop on Agrienvironmental Indicators, 2010, March. pp. 23–26. https://www.researchgate.net/publication/312595805.
- 62. Baldock J, Macdonald L, Farrell M, Welti N, Monjardino M. Nitrogen dynamics in modern cropping systems. CSIRO Agriculture and Food. 2018. https://grdc.com.au/resources-and-publications/grdc-update-papers/tab-content/grdc-update-papers/2018/02/nitrogen-dynamics-in-modern-cropping-systems.
- 63. Timsina J, Dutta S, Devkota KP, Chakraborty S, Neupane RK, Bishta S, et al. Improved nutrient management in cereals using Nutrient Expert and machine learning tools: Productivity, profitability and nutrient use efficiency. Agricultural Systems. 2021;192:103181.
- 64. Snyder CS, Bruulsema TW. Nutrient use efficiency and effectiveness in North America. Publ. Int. Plant Nutr. Inst. IPNI. 2007.
- 65. Ciampitti IA, Vyn TJ. Physiological perspectives of changes over time in maize yield dependency on nitrogen uptake and associated nitrogen efficiencies: a review. Field Crops Research. 2012;133:48–67.
- 66. Song X, Liu M, Ju X, Gao B, Su F, Chen X, et al. Nitrous oxide emissions increase exponentially when optimum nitrogen fertilizer rates are exceeded in the North China Plain. Environ Sci Technol. 2018;52(21):12504–13. pmid:30351044
- 67. Kumar D, Patel RA, Ramani VP, Rathod SV. Evaluating precision nitrogen management practices in terms of yield, nitrogen use efficiency and nitrogen loss reduction in maize crop under indian conditions. Int J Plant Prod. 2021;15(2):243–60.
- 68. Chai R, Ye X, Ma C, Wang Q, Tu R, Zhang L, et al. Greenhouse gas emissions from synthetic nitrogen manufacture and fertilization for main upland crops in China. Carbon Balance Manag. 2019;14(1):20. pmid:31889246
- 69. Hoben JP, Gehl RJ, Millar N, Grace PR, Robertson GP. Nonlinear nitrous oxide (N2O) response to nitrogen fertilizer in on‐farm corn crops of the US Midwest. Global Change Biology. 2010;17(2):1140–52.
- 70. Sapkota TB, Singh LK, Yadav AK, Khatri-Chhetri A, Jat HS, Sharma PC, et al. Identifying optimum rates of fertilizer nitrogen application to maximize economic return and minimize nitrous oxide emission from rice–wheat systems in the Indo-Gangetic Plains of India. Archives of Agronomy and Soil Science. 2020;66(14):2039–54.
- 71. Shcherbak I, Millar N, Robertson GP. Global metaanalysis of the nonlinear response of soil nitrous oxide (N2O) emissions to fertilizer nitrogen. Proc Natl Acad Sci U S A. 2014;111(25):9199–204. pmid:24927583
- 72. Menegat S, Ledo A, Tirado R. Greenhouse gas emissions from global production and use of nitrogen synthetic fertilisers in agriculture. Sci Rep. 2022;12(1):14490. pmid:36008570
- 73. Yousefi M, Damghani AM, Khoramivafa M. Energy consumption, greenhouse gas emissions and assessment of sustainability index in corn agroecosystems of Iran. Sci Total Environ. 2014;493:330–5. pmid:24951890
- 74. Kargwal R, , Kumar A, Garg MK, Chanakaewsomboon I. A review on global energy use patterns in major crop production systems. Environ Sci: Adv. 2022;1(5):662–79.
- 75. Todorović M, Mehmeti A, Cantore V. Impact of different water and nitrogen inputs on the eco-efficiency of durum wheat cultivation in Mediterranean environments. Journal of Cleaner Production. 2018;183:1276–88.
- 76. Ul Haq S, Boz I, Shahbaz P, Yıldırım Ç. Evaluating eco-efficiency and optimal levels of fertilizer use based on the social cost and social benefits in tea production. Environ Sci Pollut Res Int. 2020;27(26):33008–19. pmid:32524407