Salinization of soils and freshwater resources by natural processes and/or human activities has become an increasing issue that affects environmental services and socioeconomic relations. In addition, salinization jeopardizes agroecosystems, inducing salt stress in most cultivated plants (nutrient deficiency, pH and oxidative stress, biomass reduction), and directly affects the quality and quantity of food production. Depending on the type of salt/stress (alkaline or pH-neutral), specific approaches and solutions should be applied to ameliorate the situation on-site. Various agro-hydrotechnical (soil and water conservation, reduced tillage, mulching, rainwater harvesting, irrigation and drainage, control of seawater intrusion), biological (agroforestry, multi-cropping, cultivation of salt-resistant species, bacterial inoculation, promotion of mycorrhiza, grafting with salt-resistant rootstocks), chemical (application of organic and mineral amendments, phytohormones), bio-ecological (breeding, desalination, application of nano-based products, seed biopriming), and/or institutional solutions (salinity monitoring, integrated national and regional strategies) are very effective against salinity/salt stress and numerous other constraints. Advances in computer science (artificial intelligence, machine learning) provide rapid predictions of salinization processes from the field to the global scale, under numerous scenarios, including climate change. Thus, these results represent a comprehensive outcome and tool for a multidisciplinary approach to protect and control salinization, minimizing damages caused by salt stress.

Despite years of study and substantial investment in remediation and prevention, soil erosion continues to be a major environmental problem with regard to land use in India and elsewhere around the world. Furthermore, changing climate and/or weather patterns are exacerbating the problem. Our objective was to review past and current soil conservation programmes in India to better understand how production-, environmental-, social-, economic- and policy-related issues have affected soil and water conservation and the incentives needed to address the most critical problems. We found that to achieve success in soil and water conservation policies, institutions and operations must be co-ordinated using a holistic approach. Watershed programmes have been shown to be one of the most effective strategies for bringing socio-economic change to different parts of India. Within both dryland and rainfed areas, watershed management has quietly revolutionized agriculture by aligning various sectors through technological soil and water conservation interventions and land-use diversification. Significant results associated with various watershed-scale soil and water conservation programmes and interventions that were effective for reducing land degradation and improving productivity in different parts of the country are discussed.

Integrated management of rice–maize systems is an emerging challenge in southern India due to improper rice residues and tillage management in maize crops. Conservation agriculture (CA) practices such as a reduced tillage and maintaining stubble mulch may hold the potential to increase yields, reduce crop establishment costs and increase farm incomes. A five-year trial was performed to study the effect of different CA and establishment methods in rice on system productivity, profitability, and soil carbon status in a rice–maize system. In the rainy season, the trial consisted of two main treatments: (i) normal manual transplanting and (ii) direct-wet seeding, and three sub-main treatments at different sowing dates with fifteen day intervals. In addition, in the winter season, two tillage treatments (conventional and minimum tillage) were imposed over the rainy season treatments. Both rice and maize were grown under irrigated conditions. The results showed that sowing times at 15 day intervals did not impact the yield significantly. Transplanted rice obtained a significantly higher grain yield during the first four years, but in the last year, the yield was similar in both of the establishment methods. In the winter season, conventional tilled maize recorded a higher cob yield than under the minimum tilled treatment, except for the last year, where both the tillage treatment effects were the same. System productivity of CA-based minimum tilled rice–maize was inferior during the first three years but was superior to the conventionally tilled method in the fourth and fifth year. Pooled analysis revealed that the conventionally tilled rice–maize system resulted in a similar system productivity as that of the CA during the study period. The cost–benefit analysis revealed that transplanted rice and conventionally tilled maize fetched higher net returns of INR 111,074 and INR 101,658/ha, respectively, over the direct-wet seeded rice and CA. In addition, the 15 July rice sown followed by the maize system led to an increase in irrigation water productivity by 15.7%, and the total water (irrigation + rainfall) productivity by 27.1% in the maize crop compared with the 30 July sown system. The CA-based rice–maize system resulted in a significantly higher very labile (0.194%) and labile (0.196%) carbon concentration at a 0–5 cm depth of soil compared to those under the conventional system. Thus, CA can be recommended for southern India and similar agro-ecological tropic and sub-tropic conditions. This system can be followed with appropriate location-specific modification in South-Asian countries, where crop yields and soil health are declining as a result of continuous cereal–cereal crop rotation.

A robust forecast of rice yields is of great importance for medium-to-long-term planning and decision-making in cereal production, from regional to national level. Incorporation of spatially correlated adjacent effects in forecasting models in general, results in accurate forecast. The Space Time Autoregressive Moving Average (STARMA) is the most popular class of model in linear spatiotemporal time series modelling. However, STARMA cannot process nonlinear spatiotemporal relationships in datasets. Alternately, Time Delay Neural Network (TDNN) is a most popular machine learning algorithm to model the nonlinear pattern in data. To overcome these limitations, two-stage STARMA approach was developed to predict rice yield in some of the most intensive national rice agroecosystems in India. The Mean Absolute Percentage Errors value of proposed STARMA-II approach is lower compared to Autoregressive Moving Average (ARIMA) and STARMA model in all examined districts, while the Diebold-Mariano test confirmed that STARMA-II model is significantly different from classical approaches. The proposed STARMA-II approach is promising alternative to classical linear and nonlinear spatiotemporal time series models for estimating mixed linear and nonlinear patterns and can be advanced tool for mid-to-long-term sustainable planning and management of crop yields and patterns in agroecosystems, i.e., food supply and demand from local to regional levels.

In agroecosystems, drought is a critical climatic phenomenon that affects evapotranspiration and induces water stress in plants. The objective in this study was to characterize and forecast water stress in the Hyderabad region of India using artificial intelligence models. The monthly precipitation data for the period 1982–2021 was characterized by the standardized precipitation index (SPI) and modeled using the classical autoregressive integrated moving average (ARIMA) model and artificial intelligence (AI), i.e., artificial neural network (ANN) and support vector regression (SVR) model. The results show that on the short-term SPI3 time scale the studied region experienced extreme water deficit in 1983, 1992, 1993, 2007, 2015, and 2018, while on the mid-term SPI6 time scale, 1983, 1991, 2011, and 2016 were extremely dry. In addition, the prediction of drought at both SPI3 and SPI6 time scales by AI models outperformed the classical ARIMA models in both, training and validation data sets. Among applied models, the SVR model performed better than other models in modeling and predicting drought (confirmed by root mean square error—RMSE), while the Diebold–Mariano test confirmed that SVR output was significantly superior. A reduction in the prediction error of SVR by 48% and 32% (vs. ARIMA), and by 21% and 26% (vs. ANN) was observed in the test data sets for both SPI3 and SPI6 time scales. These results may be due to the ability of the SVR model to account for the nonlinear and complex patterns in the input data sets against the classical linear ARIMA model. These results may contribute to more sustainable and efficient management of water resources/stress in cropping systems.

Due to the introduction of zero-till wheat in the Indo-Gangetic plains (IGP) in India, irrigation and fertilizer nitrogen (N) management needs to be modified from that followed under conventionally tilled fields. A field experiment was conducted to study the effect of irrigation and N levels on yield and N uptake by zero-till wheat, fertilizer N-use efficiency, and distribution of nitrate-N (NO₃-N) in a soil profile under zero-till conditions in an acidic alluvial soil of the eastern IGP. The experiment was laid out in a split-plot design with four levels of irrigation as main plots (I₀-no irrigation, rain-fed, I₁-122 mm in one irrigation at 21 days after sowing (DAS), I₂-263 mm in two irrigations at 21 and 42 DAS, and I₃-386 mm in three irrigations at 21, 42, and 84 DAS) and 4 N levels [0 (N₀), 60 (N₁), 120 (N₂), and 150 (N₃) kg N ha⁻¹] as subplots. Grain and straw yields were significantly higher at the irrigation level-I₂ and 120 kg N ha⁻¹-N₂ over the control (I₀ and N₀) and were at par with the highest applied levels of irrigation and N (I₃ and N₃). The nitrogen uptake by wheat followed a trend similar to yield for irrigation levels; however, it increased significantly up to 150 kg N ha⁻¹. After the harvest of wheat crop, more NO₃-N was observed in the 60–90 cm subsurface soil layer than in the surface 0–15 cm and/or 15–30 cm and 30–60 cm subsurface soil layers. The highest NO₃-N concentration was recorded in the treatment I₂N₂. Accumulation of NO₃-N in the soil increased up to irrigation levels I₂ and with increasing doses of fertilizer N application. Combined applications of irrigation and N had a positive and significant influence on agronomic efficiency (AE) and apparent N recovery (ANR) but had no significant effect on physiological efficiency (PE). This study suggests that an appropriate combination of irrigation and N levels in zero-till wheat can lead to not only high-yield levels and N-use efficiency but also adequately control NO₃-N leaching under acidic alluvial soils in the eastern IGP.