Fire has traditionally been used as an easy (and cheap) method to clear land for agricultural purposes (D1 in Fig. 1; Fig. 2). This can be highly valuable to shifting cultivators, especially in developing tropical countries, where farmers are generally poor and often do not have access to heavy agricultural machinery (Bougma et al., 2025; Campos et al., 2025; Hauser and Norgrove, 2013). Therefore, the savings from avoiding more expensive mechanical methods to clear land can represent a significant economic benefit (Varma, 2003). However, as argued by Varma (2003), the economic costs of using fire can exceed the total economic benefits, so using fire for land clearing appears to be an inefficient method from a social perspective. These costs refer to the welfare loss caused by accidental fires (see below) and the negative externalities or social costs in the form of damage to valuable environmental services, such as air, water, and soil quality (Varma, 2003). We discuss these environmental costs more deeply in the “Undesired environmental impacts of fire” section.

Fig. 1.

Hypothetical desired (D, in green) and undesired (U, in red) effects of fire induced by slash-and-burn agriculture in the tropics. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).

Fig. 2.

Synthetic overview of the desired (D1–D3) and undesired (U1–U5) effects of slash-and-burn agriculture (SBA) identified in the reviewed literature. For each hypothesis, the table summarizes the overall evidence as supporting, rejecting, or non-conclusive, based on the direction of the empirical findings reported by the studies included in the review (study IDs shown). The full database of reviewed studies, including study design, ecological response variables, and geographic context, is provided in Table S1 (Supplementary Material). Asterisks indicate hypotheses for which the effect is widely recognized in the SBA context but was not directly demonstrated by the reviewed studies.

In addition to land clearing, farmers use fire to accelerate the incorporation of nutrients into the soil (D2, Fig. 1). This is a valuable outcome in the tropics, where soils are generally thin and nutrient poor. By burning plant biomass, farmers expect to release solids (ash) rich in essential nutrients for plants (e.g., N, P, K) that were previously stored in plant biomass and organic matter. If fire contributes to soil fertilization, there should be a post-fire increase in soil nutrient concentrations. Despite considerable evidence supporting this prediction (8 out of 15 studies; Fig. 2), the effect of fire on soil nutrient concentration is highly variable among studies, depending on the evaluated nutrient, fire intensity, soil type and depth, as well as the timing of assessment (Arunrat et al., 2024a; Béliveau et al., 2009). For example, Kennard and Gholz (2001) and Williams et al. (1997) found significant increases in inorganic nitrogen (NH₄+, NO₃−) in soils shortly after fire, indicating that a portion of the nutrients contained in the aboveground biomass were transferred to the soil during burning (Montagnini and Buschbacher, 1989). Higher concentrations of P, Ca, K, and Mg have also been recorded after fire (Juárez-Orozco et al., 2024). However, Kauffman et al. (1993; 1995) found losses in N and P, probably because these nutrients volatilize at temperatures well below (e.g., N ≈ 200 °C; P ≈ 300 °C) those reached by fire (> 800 °C), suggesting that fire effects largely depend on fire intensity (Kennard and Gholz, 2001). The fertilizing effect of fire also depends on soil depth, as nutrient concentration increases are limited to the soil’s surface layer (0–2.5 cm) where ash concentrates (Ando et al., 2014; Béliveau et al., 2015; Juárez-Orozco et al., 2024; Kauffman et al., 1995).

Although nutrient-rich ashes can fertilize the soil, this effect is only short-term, disappearing months after burning (Ellingson et al., 2000), when some nutrients may be depleted through leaching and surface runoff (Adeyolanu et al., 2013; Uhl and Jordan, 1984; Béliveau et al., 2017), or plant absorption (Arunachalam, 2002; Døckersmith et al., 1999). Rapid soil nutrient losses after fire due to leaching, erosion, and volatilization are well documented (Kauffman et al., 1995; Lungmuana et al., 2018; Wang et al., 2022); for example, more than 57% of the aboveground P present in the ash disappeared within just 17 days after burning, reflecting the high susceptibility of this nutrient pool to wind and water erosion in burned areas (Kauffman et al., 1993). Similarly, Giardina et al. (2000) observed that within the first month after fire, approximately 55% of P and 74% of N contained in the ash were removed from the site by wind. Therefore, the general trend is a transient nutrient enrichment followed by rapid losses. Post-fire nutrient losses can contribute to declining soil fertility in systems with shortened fallow periods, which in some regions lead farmers to burn field debris more frequently after each crop cycle and/or use chemical fertilizers. These fertilizers are often promoted by government programs, such as the “Fertilizantes para el Bienestar” (Fertilizers for Wellbeing) program in Mexico, and the “Subsidized Fertilizer Programme” in Ghana. In both countries, thousands of highly soluble chemical fertilizers are freely distributed each year, without correct monitoring of their impacts. For example, chemical fertilizers are known to pollute underground water bodies, negatively impacting aquatic ecosystems (Bijay-Singh and Craswell, 2021). However, to our knowledge, these potentially strong impacts have not yet been properly evaluated, so additional studies are needed to quantify them.

The effect of fire on soil fertilization and nutrient losses can also depend on the amount of burned biomass. If fire releases essential nutrients stored in plant biomass and organic matter, the combustion of larger quantities of plant biomass should lead to the release of a proportionally higher amount of nutrients. Thus, the older the forest that was downed and burned, the greater the amount of ash and nutrients that should be transferred to the soil, since older forests accumulate more aboveground biomass than younger forests. However, empirical evidence for this relationship is scanty, as only one of five studies supported this prediction (Kennard and Gholz, 2001), whereas most (4 studies, 80%) did not detect higher soil nutrient concentrations after burning larger biomass stocks (Andriesse and Schelhaas, 1987; Døckersmith et al., 1999; Ellingson et al., 2000; Juárez-Orozco et al., 2024; Kauffman et al., 1995). One possible explanation is that larger biomass stocks may lead to more severe fires, increasing ash-derived nutrient inputs to the soil surface while simultaneously enhancing volatilization losses and soil organic matter combustion, partially offsetting potential gains (Ando et al., 2014; Kennard and Gholz, 2001). Evidence from a slash-and-burn chronosequence further shows that nutrient increases after fire are limited to the soil surface and that nutrient concentrations decline within a few years of fallow; therefore, repeated fire-fallow cycles over decades reduce soil nutrient retention and cation exchange capacity relative to forest soils (Fachin et al., 2021). Thus, even in those situations where substantial amounts of biomass are repeatedly burned, fertilization effects remain transient and do not prevent long-term soil nutrient depletion. Therefore, the influence of burned biomass on soil fertilization is context-dependent and difficult to predict across SBA systems (Andriesse and Schelhaas, 1987; Kauffman et al., 1995).

Another important outcome of fire is the delay in the recovery of native and exotic plant species (D3, Fig. 1). Promoting such recruitment limitation is particularly valuable in the tropics, where hundreds of plant species can colonize the site and compete with crop plants for light and nutrients. However, fire effects on soil seed banks are not homogeneous and depend on fire regime characteristics (e.g., intensity and duration), soil properties and moisture, and species functional strategies (Bezerra et al., 2024). Because extreme heat can damage seed embryos and disrupt germination, fire may deplete the soil seed bank through direct seed incineration or reducing seed viability, particularly in the upper soil layers where heating is greatest. If this mechanism dominates, the number and diversity of viable seeds and/or the density of germinated seeds should be lower after fire and in burned areas compared to unburned areas, as observed in most studies that tested this prediction (Cao et al., 2000; Kennard et al., 2002; Mamede and Araújo, 2008; Miller, 1999; but see Bezerra et al., 2022; Ewel et al., 1981; Fig. 2).

When assessing fire effects across plant groups, species responses are expected to vary with seed traits and post-fire regeneration strategies. Weaker fire damage has been predicted for species with small seeds and hard coats, as these traits may confer greater heat resistance or allow seeds to persist in deeper soil layers. Many small-seeded herbs or shrubs may also benefit from increased light and reduced competition in the months following fire (Tangney et al., 2020). If this is the case, the number of light-demanding and small-seeded species should be higher after fire and in burned areas compared to unburned areas. Most studies assessing this prediction (5 of 7 studies, 71%) supported it, showing increases in herbaceous and shrubby species (Cao et al., 2000; De Rouw et al., 2014; Kennard et al., 2002; Miller, 1999; Rico-Gray and García-Franco, 1992). However, contrasting responses occur across species and sites. For example, De Rouw et al. (2014) reported losses of some light-demanding species after fire, probably because seed survival depends strongly on soil-mediated thermal gradients and burial depth. Fire heat dissipates rapidly along the soil profile, so seeds near the surface experience lethal temperatures whereas deeper seeds may remain below mortality thresholds (Cao et al., 2000). Consequently, post-fire recruitment reflects the interaction between fire intensity, soil heating profile, burial depth, and seed traits, leading to context-dependent outcomes across species and environments (Pausas and Ribeiro, 2017). Empirical evidence on the role of seed size remains inconsistent. While deeper burial could allow smaller seeds to escape lethal temperatures, Bezerra et al. (2022) detected no significant effect of fire on seed size or seed mass, whereas Mamede and Araújo (2008) reported complete mortality for very small seeds (diameter < 0.6 mm), especially monocot seeds. These contrasting results reinforce that fire effects on seed-bank structure and composition are highly contingent on local ecological conditions. Therefore, future studies should further investigate how soil properties, seed traits, and fire characteristics interact to shape the seed-bank structure and composition in SAB systems. Thus, while recruitment limitation may enhance short-term crop establishment, its cumulative effects under intensified management and shortened fallow cycles may involve ecological trade-offs, potentially constraining secondary succession and forest recovery (this undesired effect is discussed below).

Although fire releases nutrient-rich ash, it is important to note that it represents only a small fraction of the nutrients previously contained in plant biomass and soil organic matter. In fact, a large fraction of forest nutrients is lost during burning (U1, Fig. 1), as high temperatures cause elements like N, P, S and K to volatilize from the biomass pool (Kauffman et al., 1995, 1993; Wang et al., 2022; Fig. 2). For instance, Kauffman et al. (1993) found that 96% of the pre-fire aboveground N and C pools and 56% of the pre-fire aboveground P pool were lost through combustion in a tropical dry forest. This likely occurs because, prior to burning, the highest concentrations of these nutrients are found in litter, leaves, and fine wood debris, which are almost completely burned by fire, promoting nutrient volatilization. In particular, N losses exceeded 500 kg/ha and P losses exceeded 20 kg/ha during severe fires (Kauffman et al., 1993). These values are even higher in tropical rainforests such as the Amazon, where N losses ranged from 817 to 1605 kg/ha, and S losses ranged from 92 to 122 kg/ha (Kauffman et al., 1995). However, fire behavior and severity can vary widely across slash-and-burn systems, depending on vegetation type, fuel structure, litter accumulation, atmospheric conditions, and burning practices (Fachin and Thomaz, 2023). In many smallholder agricultural plots, particularly those conducted under moist conditions or with discontinuous fuels, fires tend to be of low to moderate severity (Antoneli and Thomaz, 2025). Under these conditions, a fraction of the nutrients released from burned biomass is retained locally in ash and incorporated into upper mineral soil rather than being entirely exported from the system (Thomaz et al., 2025).

Experimental evidence indicates that soil chemical responses scale with fire severity. In low-severity burns typical of SBA, most soil properties remain statistically similar to unburned controls, with only modest increases in some readily available nutrients (e.g., P and K), whereas larger increases in exchangeable base cations, particularly Ca and K, and in cation exchange capacity, occur mainly under moderate to high severity burns (Antoneli and Thomaz, 2025). Thus, although substantial nutrient losses occur from the biomass compartment, a fraction of immediate post-fire soil nutrients may be maintained or even temporarily enhanced depending on fire severity and ash retention. These apparently contrasting patterns reflect differences between biomass nutrient losses and soil-ash nutrient retention processes. Taken together, estimates of biomass nutrient losses through fire reported for tropical forests suggest that more than a century of fallow would be required for full nutrient reaccumulation, a period much longer than current average fallow times in tropical regions (< 20 yr; Schwartz et al., 2020). Consistent with this pattern, empirical evidence from African slash-and-burn systems indicates that shortening fallow periods reduces the recovery of soil nutrients and organic matter, with longer fallows showing higher soil organic carbon, total nitrogen, and exchangeable potassium contents (Kilawe et al., 2022). This indicates that once old-growth forests are converted to SBA, they are unlikely to fully regain their initial nutrient stocks under contemporary management regimes.

Related to the loss of forest nutrients, another major undesirable effect of fire is the emission of greenhouse gases (e.g., CO2 and CH4) responsible for global climate change (U2, Fig. 1). Again, the magnitude of these emissions varies considerably depending on fire intensity, biomass availability, and the land-use trajectory following cultivation. Although we could not find global estimates of SBA-related emissions, we found strong evidence supporting this undesired effect on a local scale (Fig. 2). For example, Kauffman et al. (1995) reported that in a single fire, carbon losses range from 58 to 112 Mg/ha in the Amazon, depending on the amount of aboveground biomass burned (also see Hughes et al., 2000; Kauffman et al., 2003 and Sampaio et al., 2003). Similarly, Guild et al. (1998) reported carbon losses of 79 to 102 Mg/ha, which were emitted into the atmosphere. More recent field-based measurements also show that greenhouse gas emissions increase significantly in recently burned areas, with carbon dioxide emissions being approximately twice those of unburned sites and post-fire methane emissions increasing (Dhandapani and Evers, 2020). Therefore, on a global scale, SBA-related carbon emissions may reach hundreds of megatons per year (Baccini et al., 2012). At broader spatial scales, remote sensing analyses indicate that SBA landscapes remain widespread across the humid tropics, covering hundreds of millions of hectares, highlighting the relevance of these land-use systems for regional and global carbon accounting (Heinimann et al., 2017). Although some of this carbon dioxide is captured by secondary forests once agricultural fields are abandoned (Chazdon et al., 2016; Heinrich et al., 2021), recent estimates for the Amazon, Central Africa, and Borneo suggest that the 107 megatons of CO2 absorbed by secondary forests each year only represent 26% of the CO2 emitted by the destruction of original forests (Heinrich et al., 2023). Clearly, such carbon uptake cannot fully offset the losses caused by increasing deforestation rates in the tropics (Sims et al., 2025), which in some regions may include SBA agriculture and other smallholder land-use transitions (Curtis et al., 2018).

The two major undesired effects described above can be further aggravated when fires used in SBA escape control and spread into surrounding forests (Bezerra et al., 2024; Varma, 2003; U3, Fig. 1). In fact, the Global Forest Watch platform reported that although tropical rainforest loss slowed in 2025, fire is a growing threat to forests worldwide (Goldman et al., 2026). This risk is particularly relevant in tropical regions, where burning of vegetation occurs in landscapes with high fuel load and where dry conditions may facilitate fire spread. As an example, during the 1997 dry season in Indonesia, fires associated with agricultural land clearing, coupled with the unusually dry weather linked to the ENSO climate phenomenon, ignited large forest fires that destroyed millions of hectares of land (Varma, 2003). As noted by Varma (2003), the cumulative environmental damage caused by these fires was enormous. Although there are no accurate global estimates of annual forest losses caused by accidental fires originating from SBA practices, these events can locally expand the spatial impact of agricultural burning beyond the intended cultivation area. For instance, in Yucatán, Mexico, we recorded two wildfires that spread over 200 ha that were caused by 1-ha SBA crops (J. S. Bezerra and V. Arroyo-Rodríguez, personal observation). These examples illustrate how escaped agricultural fires can amplify the impacts of slash-and-burn systems when burning is not properly controlled. However, additional studies are needed to quantify the impact of SBA on wildfires.

An immediate consequence of fire is an increase in soil temperature. In slash-and-burn systems, soil heating is primarily determined by fire severity, which is generally low to moderate but spatially variable due to heterogeneity in fuel distribution and environmental conditions. The amount of heat released and retained in the soil largely depends on fuel availability, with greater vegetation biomass generally increasing combustion intensity and prolonging heat residence time (Døckersmith et al., 1999; Kauffman et al., 1993). Experimental evidence further indicates that burns involving larger biomass loads can generate higher and longer-lasting soil temperatures, with heating effects extending deeper into the soil profile (Ando et al., 2014). However, some studies suggest that the relationship between tree biomass and soil temperature is not straightforward (Andriesse and Schelhaas, 1987; Juárez-Orozco et al., 2024). For example, Andriesse and Schelhaas (1987) observed that burning different volumes of piled biomass resulted in highly variable soil temperatures, with no consistent pattern across fuel amount. These contrasting results highlight that soil temperature responses are not solely a function of fuel load, but also reflect variability in fire severity, which in turn depends on other factors such as substrate moisture and fuel spatial arrangement, both of which determine combustion efficacy (Andriesse and Schelhaas, 1987).

In any case, soil temperatures peak at the surface, where direct contact with the burning biomass occurs, and decrease rapidly with depth due to low thermal conductivity, especially in dry and unconsolidated substrates (Andriesse and Schelhaas, 1987; Kauffman et al., 1993). For example, maximum temperatures at the surface commonly exceed 500 °C and can reach over 900 °C in high-intensity burns (Døckersmith et al., 1999; Giardina et al., 2000; Kauffman et al., 1993; Kennard et al., 2002). However, soil heating declines sharply within the first few centimeters of depth (Ewel et al., 1981), indicating that direct thermal effects are largely confined to the uppermost soil layers.

However, even when confined to the topsoil, fire-induced heating can substantially alter soil physical and chemical properties, including pH, moisture, nutrient availability, and soil organic carbon (Certini, 2005; Certini et al., 2021; Pellegrini et al., 2020). These changes are driven not only by direct heat transfer but also by combustion of organic matter, ash deposition, and volatilization losses, which together reshape the abiotic environment where soil biota persists (U4, Fig. 1). For instance, soil pH often increases immediately after burning due to ash inputs rich in basic cations (Ca, Mg, K), temporarily reducing soil acidity. This shift can favor some bacterial groups (Arunrat et al., 2024b), while reducing fungal abundance, including beneficial arbuscular mycorrhizal fungi (AMF), by altering plant-soil interactions and reducing host plant availability, leading to more bacteria-dominated communities (Barraclough and Olsson, 2018; Sharmah and Jha, 2014). In addition, fire-induced pulses of inorganic nutrients (e.g., N and P), followed by subsequent depletion, can further influence microbial activity and community composition (Arunrat et al., 2024b). As a consequence of these abiotic changes, fire can restructure soil microbial communities by favoring disturbance-tolerant taxa and reducing the abundance of groups with key ecological functions (Arunrat et al., 2024c). Nevertheless, many fire-induced changes in soil microbiota are likely to be transient, particularly under low-severity burns and in systems where fallow periods allow vegetation and soil to recover. Empirical evidence shows that the abundance and diversity of sensitive microbial groups are reduced in burned compared to unburned areas (Arunachalam, 2002; García-Oliva et al., 1999; Meyer et al., 2019; Navarrete et al., 2015; Sharmah and Jha, 2014; Fig. 2). These changes affect bacterial and fungal groups involved in key processes, such as nutrient cycling, organic matter decomposition, and soil structure maintenance. Therefore, although some microbial responses may diminish over time, recurrent burning and shortened fallow periods can impair microbial-mediated processes, potentially constraining soil recovery and vegetation regeneration in post-fire agroecosystems.

As discussed above, farmers benefit from fire in the short term by limiting unwanted plant recruitment. Nevertheless, when repeated over long time periods or combined with shortened fallow cycles, such fire-related recruitment limitation may constrain secondary succession and slow down forest recovery in abandoned lands (Arroyo‐Rodríguez et al., 2017; Barraclough and Olsson, 2018; Meyer et al., 2019; Tabarelli et al., 2025; U5 in Fig. 1; Fig. 2). Contemporary SBA systems in some regions operate with fallow periods of approximately 3–5 years, and reductions in fallow length have been widely reported (Thomaz and Rosell, 2020), potentially limiting the system capacity for full ecological recovery. Under intensified management, this may compromise the long-term sustainability of both tropical forests and the SBA itself, which depends on forest regeneration to offset the environmental degradation caused during the stages of vegetation slashing, vegetation burning, and farming (Bezerra et al., 2024; Hauser and Norgrove, 2013).

Because fire can reduce viable propagule pools (see Weed control section), forest recovery in SBA-dominated landscapes becomes increasingly dependent on seed rain and resprouting strategies (Arroyo‐Rodríguez et al., 2017; Poorter et al., 2024). In tropical dry forests, however, a large proportion of seeds arriving via seed rain may be non-viable due to harsh conditions that damage embryos, especially in small, recalcitrant, and animal-dispersed seeds lacking dormancy (Bezerra et al., 2023a, b). Therefore, regeneration tends to be dominated by disturbance-adapted plant species, especially those with abiotic dispersal syndromes, hard-coated seeds, or high resprouting ability (Miller and Kauffman, 1998; Rico-Gray and García-Franco, 1992; Vanderlei et al., 2024). Similar patterns have been reported in African SBA systems, where shortened fallow periods were associated with reduced tree species richness and recovery, as well as increased dominance of shrubs and disturbance-adapted vegetation during succession (Kilawe et al., 2022). Fire-induced topkill often stimulates the activation of dormant buds, increasing resprouting incidence and altering forest structure and composition during secondary succession (Barros et al., 2021; Vanderlei et al., 2024). However, resprouting ability varies among species and may depend on fire severity and pre-disturbance carbohydrate reserves stored in stems and roots (Baião et al., 2024; Kennard et al., 2002; Miller and Kauffman, 1998). Under recurrent burning and shortened fallow cycles, these trait-mediated filters may progressively simplify forest structure and redirect successional trajectories. Further research is needed to clarify how fire severity, propagule limitation, carbohydrate storage, and species-specific traits interact to shape long-term regeneration success and successional pathways across heterogeneous SBA systems.