Agroforestry systems (AFS) and practices followed in India are highly diverse due to varied climatic conditions ranging from temperate to humid tropics. The estimated area under AFS in India is 13.75 million ha with the highest concentration being in the states of Uttar Pradesh (1.86 million ha), followed by Maharashtra (1.61 million ha), Rajasthan (1.55 million ha) and Andhra Pradesh (1.17 million ha). There are many forms of agroforestry practice in India ranging from intensified simple systems of monoculture, such as block plantations and boundary planting, to far more diverse and complex systems, such as home gardens. As a result, the biomass production and carbon sequestration potential of AFS are highly variable across different agro-climatic zones of India. Studies pertaining to the assessment of biomass and carbon storage in different agroforestry systems in the Indian sub-continent are scanty and most of these studies have reported region and system specific carbon stocks. However, while biomass and carbon stock data from different AFS at national scale has been scanty hitherto, such information is essential for national accounting, reporting of C sinks and sources, as well as for realizing the benefits of carbon credit to farmers engaged in tree-based production activities. Therefore, the objective of this study was to collate and synthesize the existing information on biomass carbon and SOC stocks associated with agroforestry practices across agro-climatic zones of India. The results revealed considerable variation in biomass and carbon stocks among AFS, as well as between different agro-climatic zones. Higher total biomass (>200 Mg ha⁻¹) was observed in the humid tropics of India which are prevalent in southern and northeastern regions, while lower total biomass (<50 Mg ha⁻¹) was reported from Indo-Gangetic, western and central India. Total biomass carbon varied in the range of 1.84 to 131 Mg ha⁻¹ in the agrihorticulture systems of western and central India and the coffee agroforests of southern peninsular India. Similarly, soil organic carbon (SOC) ranged between 12.26–170.43 Mg ha⁻¹, with the highest SOC in the coffee agroforests of southern India and the lowest in the agrisilviculture systems of western India. The AFS which recorded relatively higher SOC included plantation crop-based practices of southern, eastern and northeastern India, followed by the agrihorticulture and agrisilviculture systems of the northern Himalayas. The meta-analysis indicated that the growth and nature of different agroforestry tree species is the key factor affecting the carbon storage capacity of an agroforestry system. The baseline data obtained across various regions could be useful for devising policies on carbon trading or financing for agroforestry.

Land use of the sub-Himalayan region is not that intensive like the intensively land-managed region of Punjab, India. Land resources of the sub-Himalayas must be managed effectively for sustainable development by preparing carbon inventories and data banks. Such macro-level studies have not been conducted yet in the present study area, and thus were conducted to suggest sustainable land use management options. To achieve the present study’s desired goal, 33 tree-based land uses were identified from forested and agricultural landscapes of the sub-humid tropical region of West Bengal, India. Stratified random nested quadrat sampling was adopted for the study. The SOC, biomass, and carbon accumulation significantly differed. Mixed forests had the highest soil primary nutrients and carbon stock. Positive correlations were observed between SOC, total standing biomass, litter production, and ecosystem carbon. The sequence of land uses based on carbon stock was mixed-species forest > sole tree species stands in a forest landscape > tea plantations > homegardens. This baseline information can be used for developing prediction models for future interventions towards sustainable land management. The study, however, could not estimate the carbon fluxes in and out of the systems due to the absence of detailed land use land-cover databases.

Land resources have been under tremendous anthropogenic pressure with the consequence of their degradation. It is therefore necessary that the land resources must be managed effectively for sustainable development. Different from the developed countries, carbon inventories and data bank to monitor carbon sequestration potential of different ecosystems are unavailable in India. Micro-level studies are essential for sustainable land use management for a land scarce nation like India. To achieve the desirable goal of the present study, a total of 33 tree-based land uses were identified from forested and agricultural landscapes. Of these total land uses, five were in forest landscapes and rest in agricultural landscapes categorized into forest tree plantations (8 land uses), agroforestry (nine land uses), commercial crop plantations (six land uses) and fruit orchards (five land uses). A stratified random nested quadrate sampling method was adopted for vegetation analysis of the different land uses. The SOC, biomass and carbon accumulation in the tree-based land uses were significantly different from each other. Mixed forest soil had the highest amount of SOC, primary nutrients, standing biomass carbon, and ecosystem carbon. Positive correlations were observed between SOC, total standing biomass, litter production, and ecosystem carbon. The sequence of best tree based land uses in terms of total SOC (up to 60 cm depth), total plant biomass, total plant biomass carbon and ecosystem carbon was mixed species forest (126.67, 781.21, 390.61 and 517.27) > sole tree species stands in forest landscape (109.71, 192.56, 96.28 and 205.98) > tea plantations (103.19, 77.07, 38.54 and 141.74) > homegardens (90.34, 97.38, 48.69 and 139.02) > mixed plantation of Anthocephalus cadamba + Swietenia macrophylla (60.07, 111.86, 55.93 and 116.02) > Swietenia macrophylla based agroforestry (62.49, 83.82, 41.91 and 104.40) > mixed plantation of Tectona grandis + Milvus migrans (60.0, 85.97, 42.99 and 102.90). Similarly, the order of the major land uses was forest > commercial crop plantation > forest tree plantations > agroforestry > fruit orchards. The overall average ecosystem carbon accumulation in forests was 3.24 times more than the land uses in agricultural landscapes. The ecosystem carbon accumulation in the tree-based land uses in both forest and agricultural landscape was highly variable and was significantly different from each other. Land use conversion from forest to agriculture can reduce more than half of the carbon stock, but converting into homegardens, tree plantations or agroforestry enhanced carbon storage of the land use systems. The present findings can be used as baseline information for developing prediction models for probable effects of different land use, future intervention and sustainable management of land use systems.

Soil biodiversity is a key indicator of soil health and soil fertility. Earlier studies focused on ecological and taxonomic aspects and documentation of soil faunal diversity across Indian agroecosystems. However, the soil-crop-biodiversity-regional nexus was limitedly explored. In the present study, we synthesised 41 studies published from 1970 to 2025 to identify the (a) research gap and opportunities, and (2) assess the relationship between agroecosystem and soil biota using the ROSES protocol. A total of 205 soil faunal species/orders, including microfauna (47), mesofauna (56), and macrofauna (102) respectively, from 11 agricultural and tree-based land uses practised in eight states and two union territories were recorded. A total of 20 soil faunal groups across six soil types or orders were recorded in Tripura and Tamil Nadu, while a weaker association between soil faunal diversity and soil type was observed in Leh and Ladakh, Meghalaya, and Rajasthan. Synthesis revealed that Inceptisols and sandy loam soils support the maximum diversity of ants, earthworms, and microarthropods, particularly Collembola and Cryptostigmata. Soil faunal diversity across the states ranged between 0.3 and 3.8, respectively, in Uttar Pradesh and Meghalaya. Croplands under conventional farming with high levels of inorganic fertiliser and pesticides support lower soil faunal diversity and abundance. Isolates and mesofauna were strongly correlated (r = 0.98), while species were negatively correlated with both mesofauna (r = -0.17) and isolates (r = -0.28). The principal component analysis suggested that most agroecosystems share similar soil faunal communities, whereas mixed plantations and mango orchards harbour unique, divergent communities. The nexus among soil microbial geomics, soil micro-foodweb, soil type, and crop-soil-nutrient dynamics needs to be explored to understand the significance of soil biodiversity in maintaining soil health. The synthesis recommends that Indian soil biodiversity work should align with the goals of the Global Soil Biodiversity Atlas (GSBA) and the Global Soil Biodiversity Initiative (GSBI), considering soil biodiversity a primary component of nature-based solutions (NbS) for sustainable agriculture.

The integration of disease severity with real-time meteorological variables and advanced machine learning techniques has given valuable predictive insights for assessing disease severity in wheat. This study emphasizes the potential of machine learning models, particularly artificial neural networks (ANN), in predicting wheat disease severity with high accuracy. Field experiment was conducted over two consecutive rabi growing seasons (2023 and 2024), using randomized block design with four sowing dates to investigate critical weather-disease relations for two key wheat pathogens: Puccinia striiformis f. sp. tritici (yellow rust) and Blumeria graminis f. sp. tritici (powdery mildew). Weekly assessments of disease severity were combined with meteorological data and analyzed using ANN and regularized regression models. The ANN model demonstrated superior predictive accuracy for both yellow rust and powdery mildew, achieving R-squared value (R2) of 0.96 and 0.98 for calibration, and 0.93 and 0.95 for validation, respectively. Random Forest (RF) models also exhibited robust performance with R2 values of 0.97 and 0.98 for calibration and 0.93 and 0.90 for validation for yellow rust and powdery mildew respectively, while Elastic Net, Lasso, and Ridge regression models showed comparatively moderate predictive capabilities. Principal component analysis (PCA) explained the key meteorological variables influencing disease incidence, with evapotranspiration, temperature, wind speed and humidity emerging as critical factors. Disease prediction is an important aspect to develop decision support system and make farmers to take informed decision to optimize production.