Journal of Experimental Agriculture International

Pesticide resistance in tropical agriculture is often treated as a technical defect that can be corrected by replacing a failing product. This interpretation overlooks the chemical treadmill: an eco-evolutionary feedback in which repeated exposure selects resistant genotypes, control failure encourages higher doses or faster product substitution, and simplified production systems preserve the selection regime. The problem is especially consequential in oilseed and cash crops because cotton, soybean, groundnut, Brassica oilseeds and plantation crops commonly connect long growing seasons, overlapping hosts, mobile pest populations and strong commercial incentives for preventive treatment. This critical narrative review evaluates resistance evolution and management through Indian evidence and global comparators, with emphasis on insecticides, transgenic insecticidal traits, herbicides and fungicides. Literature published from 1980 to 5 June 2026 was selected through accessible scholarly indexes, digital object identifier records, official article pages and citation chaining, while foundational earlier studies were retained where conceptually necessary. Evidence is strongest for repeated field-evolved resistance in Indian cotton pests, including Helicoverpa armigera, Pectinophora gossypiella, Spodoptera litura and Bemisia tabaci. Comparative cases involving Bacillus thuringiensis (Bt) cotton in Arizona, glyphosate-resistant Amaranthus palmeri in cotton-soybean systems, soybean rust in Brazil and endosulfan-resistant coffee berry borer in New Caledonia show that different pesticide technologies can enter the same treadmill when exposure is spatially continuous and governance is fragmented. Resistance is generated by interacting target-site, metabolic, penetration, behavioural and genomic mechanisms, but its operational expression depends equally on refuges, crop calendars, product markets, diagnostics and collective action. Rotations and mixtures can delay failure only when their biological assumptions are satisfied; product substitution alone usually redistributes rather than removes selection. Durable management therefore requires coordinated reduction of exposure, diversification of mortality, preservation of susceptible populations, area-wide host and residue management, and surveillance that links phenotype to mechanism and field performance. The review concludes that resistance should be governed as a shared landscape resource problem rather than managed solely as an individual farm input decision.