Search

The problem of very high-accuracy classification has led to the development of several methods, systems like – XGBoost which often get state-of-the-art on tabular data classification and regression tasks. Deep learning methods on the other hand fail to achieve high accuracy most of the time where – 1. The dataset size is restricted 2. The tabular format of the dataset. We demonstrate custom models that get generated from AutoML systems and automatically tuned for peak hyperparameter settings, perform competitively well, with no human intervention for model creation, hyperparameter tuning, and dataset manipulation tasks to achieve state-of-the-art results compared to other methods and systems out there. We briefly discuss the approaches in AutoML to construct, train and tune deep learning models automatically and discuss the results we got from it on a hackathon competition organized by NCPI.

A study to assess the profile distribution of important soil attributes in Alfisols and Entisols of West Bengal was conducted during 2016-17. Purposefully selected random sampling was carried out to collect the soils from different locations of two study sites, viz., Kalinagar (25º27'33.9"N, 88º19'10.2"E) from Malda district and Durganagar (26º09'62.7"N, 89º53'51.7"E) from Cooch Behar district of West Bengal at 0-15, 15-30, 30-45 and 45-60 cm depths. Understanding of vertical distribution of soil fertility indicators like soil organic carbon (SOC), total nitrogen (TN) and other important properties in two different soil and climatic conditions will provide an insight regarding the behaviour of soil with the change in environmental conditions. Soil bulk density (BD), porosity, pH, SOC, TN, C:N ratio and texture were determined using standard laboratory procedures and computations. Obtained results were subjected to statistical analyses. Soils of Kalinagar sites were slightly acidic in nature while soils of Durganagar were neutral in nature. Kalinagar soils were silt clay loam in texture where Durganagar soils classified as loam to sandy loam. Soil BD values increased with depth in both Kalinagar (Alfisol) and Durganagar (Entisol). The porosity percentage progressively decreased with an increase in depth. Soils of Durganagar reported higher soil porosity at all the depths studied. An increase in soil pH with increasing depth was observed in both the sites. The mean total organic carbon (TOC) content recorded maximum in surface soil and its concentration decreased with the depth. Kalinagar soils observed 7.63% higher TOC (17.94 g kg-1) content than Durganagar (16.57 g kg-1) at surface depth (0-15 cm) and its accumulation at the lower depths was also maximum in former soil. Mean TN values were also found to decrease by increasing the depth. The accumulation of total nitrogen at the subsequent depths was relatively higher in Kalinagar than Durganagar. Increase in C:N ratio with increasing depth was noticed in Kalinagar site but the opposite trend was accorded in case of Durganagar. Accumulation of SOC and TN throughout the soil depth was found to be greater in Alfisol (Kalingar) due to higher clay and silt fractions as compared to Entisol (Durganagar). There was a significant positive relation of TOC with clay and silt (r = 0.285, p

Cereal crop production gains under conventional agricultural systems in India have been declining in recent years because of inadequate management practices, creating a considerable concern. These activities were shown to deplete soil organic matter stocks, resulting in a decrease in microbial activity and soil organic carbon (SOC) content. Moreover, even with minimal use of organic sources in cultivated land, soil carbon status deteriorated, particularly in subtropical climates. Integrated nutrient management (INM), a modified farming method, has the potential to effectively utilize organic and inorganic resources, to improve the quality of soils and crops, and making farming more economically viable and sustainable. The aim of this study was to use meta-analysis to quantify the effects of INM on crop production, soil carbon, and microbial activity in Indian soils. During the years 1989–2018, data from various research studies in India, mainly on nutrient management in rice and wheat crops, were collected. Meta-Win 2.1 software was used to analyze the results, and significance was determined at p < 0.05. The results showed that the yield of rice and wheat was 1.4 and 4.9% more in INM than that in 100% NPK (N: nitrogen, P: phosphorous, and K: potassium), and that respective yields were comparatively higher in loamy soils (2.8%) and clayey soils (1.0%). The INM treatment increased SOC and microbial biomass carbon (MBC), resulting in improved overall soil quality. The SOC stock was increased by 23.8% in rice, 15.1% in wheat, 25.3% in loamy soils, and 14.4% in clayey soils in INM over 100% NPK. Microbial quotient (MQ) data showed significant trends between different management systems in both soil types, for example, INM > 100% NPK > No NPK. Due to more soil cracking and reduced aggregate stability in the rice field (greater short-term soil structural changes), the SOC stock loss in rice was higher than that in wheat. The CO2 equivalent emissions were 7.9 Mg ha−1 higher in no NPK (control) than in 100% NPK, and 16.4 Mg ha−1 higher in control than in INM. In other words, INM increased soil carbon sequestration by 2.3 Mg ha−1 as compared to using 100% NPK. Overall, the findings of this study show that INM could be a viable farming system mode in India for improving crop production, increasing soil carbon sequestration, and improving microbial activity while remaining economically and environmentally sustainable.

Mineralization of carbon (C) is a burning issue that is regulated by soil attributes. It has direct impacts on crop productivity and quantification of organic residues addition in soil. For better understanding and achievement of potential tillage benefits, a comprehensive scientific understanding of C mineralization is very important. Therefore, a laboratory incubation experiment was conducted to investigate the C mineralization rates and kinetics of crop residues (rice and maize) when applied on the surface (as zero-tillage, ZT) and incorporation (as conventional tillage, CT) in four different soil types (S1 and S2 of Entisol, S3 and S4 of Inceptisols) of West Bengal state, India. Results showed that after 7 days of incubation, there was a rapid phase of decrease in CO2-C. It continued up to day 14 followed by a sluggish nature of CO2 emission up to day-42, and after that almost levelling off in all subsequent periods up to the end of 126 days of incubation. It was evident from the kinetic models that C mineralization from the residues followed the exponential model: C = C0 (1 − e−kt). Similar rate constant (k) values were recorded in both placement methods, but the rate of maximum potential mineralizable (C0k) residue C was higher under residue incorporation treatments for both rice and maize residue. However, the rice and maize residues showed almost similar amounts of C mineralized over time when applied on the surface. The future prediction analysis using the equation C = C0 × e−kt suggested that the residues incorporated into the soil release a maximum C irrespective of residue type. We conclude that the residues when incorporated into the soil significantly increase the C footprints through maximum C mineralization, leaving the crop residues on the soil surface reduces the C footprints which helps in achieving sustainability from an environmental perspective.

A study on nutritional and microbial analysis of Kunapajala with different storage time interval was conducted in the Department of Soil Science & Agricultural Chemistry and the Department of Plant Pathology, UBKV, Coochbehar-736165, West Bengal during March, 2019. The motive of this work was to estimate the physicochemical properties, macro and micro nutrient content and various microbial load of Kunapajala with different storage time interval. Kunapajala had the highest P, K, Ca, Mg, Fe, Zn, Cu & Mn 40 days after preparation and it had highest N and S 20 days after preparation. It had the highest beneficial microbial load of Fungi, Actinomycetes, Pseudomonus, Phosphorus Solubilising Bacteria (PSB), Azotobacter, Azospirillum, Rhizobium and Trichoderma 40 days after preparation. So, continuous foliar and soil application of Kunapajala from 20 days after preparation to 40 days after preparation was beneficial to get maximum utilization. Moreover, Kunapajala can be used as an alternative against chemical fertilizers and pesticides to develop organic farming.

The fundamental role of soil organic carbon (SOC) in maintaining soil quality and regulating Earth’s carbon cycle generates renewed interests of scientists to track its status in the ecosystem. Changes in SOC dynamics is a resultant of soil’s complex interactions with vegetation, climate, and land-use practices. Even a small variation in SOC content could bring a significant impact on the atmospheric concentration of carbon dioxide. The fluxes of carbon (C) in soil are mainly dominated by its different pools rather than the total SOC. Therefore, profiling of SOC pools is very important for developing sound management practices that can sustain crop productivity and soil fertility and also reduce C emissions or mitigate climate change. However, accuracy, time requirement, and cost-effectiveness of the present analytical methods direct the need for advancing towards the standard protocols. A systematic appraisal and critical investigation of the different SOC determination methods applied by diverse research groups in contrasting agroecosystems and management conditions will improve our understanding and bridge the gaps in selecting the right protocols.

A 12-week laboratory incubation experiment was conducted to determine the effect of chemical composition of residues of nine plant species on their cumulative net nitrogen (N) mineralization in soil under submergence (SUB) and field capacity (FC) water conditions. The influence of residues and of water regimes on residue-N released over time was significant. Net mineralization was higher in FC than SUB especially in later periods, although differences either among soil water levels or among species were not significant in all the periods sampled. Net N release had significant correlations with initial N and ratios of carbon, lignin or polyphenol or the (lignin + polyphenol) to N concentration of residues under two water conditions. (Lignin + polyphenol):N ratio in SUB and carbon/nitrogen (C:N) ratio in FC showed highest correlations with N mineralization in all periods and explained 64.1% variation in N mineralization from residues under SUB and 64.5% under FC after 2 and 8 weeks, respectively.

In the Eastern Gangetic Plains of South Asia current agronomic practices are energy-intensive, and this contributes to higher cropping system production costs and CO2-equivalent greenhouse gas emissions. Increasingly uncertain and scarce resources, together with an increasingly variable and changing climate, make conventional crop production vulnerable and unsustainable. We hypothesized that replacing traditional crop management with conservation agriculture based-sustainable intensification practices would reduce energy requirements and CO2-equivalent emissions while increasing energy use efficiency. Across three major cropping systems (rice-wheat, rice-maize and rice-lentil) we tested improved management practices against a conventional baseline on over 400 on-farm trials. Improved management practices significantly reduced the total energy used in rice-wheat cropping systems, from 30,045 MJ ha−1 under traditional baseline practice to around 26,500 MJ ha−1. Similar savings in total energy used were observed in rice-maize and rice-lentil cropping systems; additionally, energy use was significantly more efficient under improved management in all three cropping systems. CO2-equivalent emissions were reduced by around 10% under cropping systems with improved management compared to emissions from baseline systems. For all cropping systems the rice-equivalent yield and net income were inversely proportional to the specific energy of the system under improved management but not under traditional practice. This study demonstrated that high cropping system yields under improved management practices increased energy use efficiency while reducing CO2-equivalent emissions. Our research enables farmers in the Eastern Gangetic Plains to produce more food, more sustainably, under increasing climate variability and change. These results are applicable across the Eastern Gangetic Plains region of over 450 million people and also have broader application for smallholder cropping systems in developing and emerging-economy countries globally.

This study aims to assess the effect of cropping systems on depth-wise distribution of important soil properties like bulk density (BD), pH, porosity, total organic carbon (TOC), total nitrogen (TN), C:N ratio, and texture. Purposeful random sampling was carried in the farmer’s fields under two cropping systems, viz., rice-maize (RM) and rice-jute (RJ) on Entisol of Durganagar (26°09’62.7"N, 89°53’51.7"E) village under Dinhata district of West Bengal (India). Soil samples were collected at four soil depths (0- 0.15, 0.15-0.30, 0.30-0.45, and 0.45-0.60 m). Results suggested that cropping systems and soil depths significantly influenced the physicochemical properties of soil. The RJ system showed significantly (p < 0.05) higher TOC and TN contents as compared to the RM system. Increased TOC in the RJ system lowered the BD and enhanced the porosity percentage of the soil. The loamy texture in the RJ system resulted in more nutrient accretion capacity than the RM system which was sandy loam in texture. The TOC and TN contents were in positive correlation with clay and silt particles but showed a negative relationship with sand. High sand content in the RM system recorded lower values of soil pH due to leaching losses of basic cations or salts. Soil pH and BD increased with depth but a reverse trend was noted in porosity percentage, TOC, TN, and C:N ratio.

Worldwide the increasing use of inorganic nitrogenous fertilizer becomes a threat to the environment and consequently to the mankind. It is high time to think about the alternate nitrogenous source for assuring sustainable agriculture. Biological nitrogen fixation (BNF), involving beneficial microorganism, is a process of converting the free atmospheric nitrogen into plant available form. This biological process harmonizes the ecosystem and offers an economic and environment-friendly approach for reducing the external inputs and improving internal sources. In nature a symbiotic relationship exists between most of the agriculturally important leguminous plants and beneficial microorganism, where bacteria can fix atmospheric free N2 and provide it to the plant in available form in exchange of nutrition and shelter. These rhizobia (bacteria) dwell in the nodule present in the roots of leguminous plants. This type of symbiosis also exists between free-living microorganisms, viz., Azolla and Anabaena, and with rice plants in anaerobic condition. Therefore, BNF is considered as an important biological process for harnessing soil health as well as for assuring economic, environmental, and agronomic benefit. However, the necessity of inclusion of leguminous crop in the cropping system due to BNF concern is not getting popularize among the farming community particularly to the marginal and low-income group farmer due to lack of availability of specific rhizobia strain for specific crop, as well as socioecomic constraints. The present book chapter is focusing on the importance of BNF in agricultural system and the effectiveness of various legume species and their beneficial microsymbiont. The genetics, biochemistry behind the BNF, and the probable strategy for improving the N2 fixation process are also getting concern for understanding these important biological phenomena.

Agriculture is facing constant challenges with a growing population and climate change. The sector is very fragile and risk-prone. In order to maintain sustainability in crop production, we must be prepared to adopt all climatesmart technologies that can reduce the adverse effects of a warming climate. There is an urgent need to assess climate change impacts and frame policies related to climate change mitigation. Modern agriculture must include the Sustainable Development Goals of the United Nations so that we not only achieve food and nutritional security but also confirm the livelihood security of our nation. This article is an overview of the state of changing climate and our management practices tackling every challenge that may encounter in the development process.

A laboratory experiment was conducted over 12 weeks to study the nitrogen (N) mineralization pattern and kinetics from nine locally available green manure (GM) at submergence and field capacity (FC), when incorporated into typical rice growing teesta alluvial soils of the district Coochbehar located at Terai Agro-climatic region of West Bengal. Rapid mineralization was evidenced by the fact that out of total N mineralized at the end of incubation, 20.43–100.59% and 27.60–76.17% released during 1st week at submergence and FC, respectively. After 12 weeks highest N was mineralized from C. sophera (162.57 mg kg−1 soil) and G. sepium (195.72 mg kg−1 soil) amended soil at submerge and FC moisture condition, respectively. At submergence, soils amended with I. cornea and V. ungiculata and at FC, G. sepium and E. indica amended soil exhibited most consistent trend of successive increase till the end with a few exception; whereas mean (of moisture regime) cumulative mineralization showed most consistency in case of V. ungiculata and C. sophera. Mean (of GM) cumulative mineralization was better at submergence over FC moisture regime during early period; whereas reverse was the trend during later period of incubation. Similar was the trend for individual GMs except C. sophera, G. sepium and I. cornea recording higher value at FC almost all throughout the incubation period. FC moisture regime recorded higher mineralization potential (N0) and lower rate constant (k) for all GMs; except N0 value for P. pinnata. Considering all these above C. sophera and V. ungiculata were identified as best GM species across moisture regime as well as at submergence; whereas at FC moisture regime G. sepium and E. indica were best performing GM species.