Over the last three decades, few studies have been conducted to tackle the complexity and heterogeneity of Bangladesh farming systems. We address these research gaps with a new survey. Accordingly, a survey was conducted in North‐Western Bangladesh to understand how socio‐economic traits influence technology adoption and to identify and characterize key farm types. The survey was based on farm household characteristics, farm structure, farming practices and livestock as well as the economic performance of the farm. Principal component analysis (PCA) and cluster analysis (CA) were used to establish the different farm typologies, and the data set based on 27 variables was carefully analysed. The findings confirmed that the key variables that significantly affect the adoption of new agricultural technologies relate to age, farming experience, level of education of the household head, income, access to markets, land ownership, the proportion of hired labour, savings, food self‐sufficiency and income from off‐farm activities. Four main farm types were identified in the study area based on resource endowment and livelihood orientation. These are (1) well‐resourced farmers entirely dependent on agriculture and less reliant on off‐farm activities; (2) moderately resourced households, which are headed by an older male with greater farming experience and which are engaged in both on‐farm and off‐farm activities; (3) resource‐constrained households with cattle as the main livestock and with income generated by the sale of livestock products; and (4) severely resource‐constrained households which are headed by young farmers/men and where income is generated by off‐farm activities. These four farm categories represent the heterogeneity of farms in North‐West Bangladesh, and it is hoped that the development of this farm household typology will help particularly the extension service, to set up appropriate extension advice that will benefit the farming community.

The rice-maize (R-M) system is rapidly expanding in Bangladesh due to its greater suitability for diverse soil types and environments. The present conventional method of cultivating puddled transplanted rice and maize is input-intensive, decreases soil health through intense ploughing, and ultimately reduces farm profitability. There is a need to investigate alternatives. Accordingly, we conducted a replicated 2-year (2020–2021) field study to investigate the effects of conservation agriculture (CA) based tillage and crop establishment (TCE) techniques and residue management practices on the physical, chemical, and biological properties of soil along with crop productivity and the profitability of rice-maize systems in the sandy loam soil of Northwest Bangladesh. Two TCE techniques Puddled transplanted rice (PTR) followed by Conventional tillage maize (CTM) and strip tillage direct-seeded rice (STDSR) followed by strip-tilled maize (STM) were assigned to the main plots and different percentages of crop residue retention (0, 25, and 50% by height) were allocated to the subplots. Results showed that a reduction in bulk density (BD), soil penetration resistance (SPR), and increased soil porosity were associated with STDSR/STM-based scenarios (strip tillage coupled with 25 and 50% residue retention). The soil organic carbon (SOC) fractions, such as dissolved organic C (DOC), light and heavy particulate organic matter C (POM-C), MAOM, and microbial biomass C (MBC) levels in the 0–10 cm layer under ST based treatments were 95, 8, 6, 2 and 45% greater, respectively, compared to CT with no residue treatment. When compared to the CT treatment, the DOC, light POM-C, heavy POM-C, and MAOM in the 10–20 cm layer with ST treatment were 8, 34, 25, 4 and 37% higher, respectively. Residue retention in ST increased average rice, maize, and system yields by 9.2, 14.0, and 14.12%, respectively, when compared to CT. The system gross margin and benefit-cost ratio (BCR) were $1,515 ha−1 and 1.90 under conventional tillage to $1,696 ha−1 and 2.15 under strip-tillage practices. Thus, our study suggests that CA could be an appropriate practice for sustaining soil fertility and crop yield under R-M systems in light-textured soils or other similar soils in Bangladesh.