Journal of Geography, Environment and Earth Science International

Human-ignited fires now dominate the fire regimes of many tropical and subtropical forests, woodlands and savannas, including the landscapes where rural households gather fruits, leaves, grasses, fungi, resins and medicinal plants for subsistence and income. Two literatures have grown largely in parallel: one on the effects of fire on plant diversity and regeneration, and another on the ecological sustainability of harvesting non-timber forest products. This critical narrative review brings them together to examine how anthropogenic fire affects floral diversity and the resource base of non-timber forest products, to explain why published findings disagree, and to identify defensible priorities for research and management. Literature published between January 1990 and July 2026 was identified through searches of multidisciplinary scholarly indexes, an agricultural bibliographic database, institutional sources and backward and forward citation tracking, and was appraised for design, replication, scale and relevance. The evidence converges on a conditional conclusion. Fire frequency, rather than the presence of fire per se, is the attribute most consistently associated with change: frequent dry-season burning in seasonally dry and humid tropical forests tends to reduce tree diversity, suppress sapling recruitment and favour fire-tolerant resprouters, whereas low or intermediate frequencies are often associated with equal or higher diversity, particularly in the herbaceous layer and in savanna-like systems. Effects on non-timber forest products are product-specific and frequently opposite in sign over different time horizons. Burning can raise short-term yields of grasses, essential-oil crops and some fungi while reducing recruitment of fruit-bearing trees and interacting with harvest, grazing and plant invasions in ways that single-factor studies cannot detect. Confidence is limited by the predominance of short-term, space-for-time and single-site studies, inconsistent definitions of fire regime attributes and scarce demographic data for most traded species. The review argues that management should move from blanket fire suppression towards regime-specific, co-designed fire management that accounts for product life history and interacting disturbances, and it proposes long-term demographic, experimental and participatory research to test this approach.