Natural or anthropogenic environmental disturbances are key events in terrestrial and aquatic systems that may influence the structuring of populations, communities, and ecosystems (Mouillot et al., 2013). Their effects on communities can be seen across different temporal and spatial scales (Turner, 2010). The type, frequency, and intensity of disturbances determine the extent to which communities are affected and their ecological resilience following each disturbance (Tanentzap et al., 2014). Disturbance events can act as extinction filters (Drapeau et al., 2016), and the type and extent of the disturbance can be of great importance for the strength of these filters (Betts et al., 2019). The resistance and resilience of communities to disturbances can vary among taxonomic groups and the local environmental conditions, reinforcing the context-dependency of biological responses to disturbances (Garnier et al., 2017). Despite the importance of disturbances for community organization and the variety of biological responses, most studies are limited to immediate temporal effects and focus on determining the effects of disturbances in isolation, rather than considering the interactive effects of multiple disturbance types (Kéfi et al., 2019).

The effects of different disturbances on biological communities can be classified as dominant, multiplicative, or additive when evaluated independently and as synergistic or antagonistic when considered interactively (Côté et al., 2016). If more than one disturbance impacts the community and the total result of these disturbances is the same as the sum when individually measured, the effect of the disturbances is classified as ‘additive’ (Côté et al., 2016). In contrast, synergistic disturbances have a greater combined effect, and antagonistic disturbances produce a weaker combined effect than if those disturbances occur as isolated incidents (Folt et al., 1999).

Most disturbances with the potential to change biological patterns do not occur in isolation (Graham et al., 2011). Despite this, some disturbances are more often studied as isolated events, such as floods (Oliveira et al., 2014) and fires (Karavani et al., 2018). These disturbances are often associated with the expansion of anthropogenic activities, such as increases in human induced fires, and climate change. Thus, studies evaluating the response of biological communities to the combination of these disturbances are necessary to understand the full effects of anthropogenic change on our ecosystems (Oliveira et al., 2014; Paolucci et al., 2016).

Floodplains are riverbank ecosystems subject to periodic flooding during the rainy season (Tomas et al., 2019). They have high productivity and rich biodiversity that support vital ecosystem services locally, regionally, and globally, such as ecotourism, river flooding mitigation, and habitat for rare and endangered species (Mitsch et al., 2015). Previous studies have shown that the period for which these areas are flooded, a possible indicator of the disturbance intensity, directly impacts terrestrial communities (Poff et al., 2018; Tanentzap et al., 2014). Rainfall has an important effect on floodplain communities, yet there is growing evidence in the literature that other disturbances, such as fire, also have significant effects on these ecosystems (Kleindl et al., 2015). Fires are becoming more frequent over time, especially those of anthropic origin (Thielen et al., 2020).

In the Pantanal biome, one of the largest freshwater floodplains in the world (Harris et al., 2005), floods are constant disturbances, shaping the vegetation structure and their organisms for over 2.5 Ma (Assine and Soares, 2004). The Pantanal is highly seasonal, and during the wet season, a vast amount of its plains are covered with water, while most of these plains dry completely during the dry season, when most fire events occur (Nunes da Cunha et al., 2006). While natural fires occur in the Pantanal before human occupation (Power et al., 2016), cattle ranchers have historically used them for pasture management at small spatial scales (Junk et al., 2006). Recently, however, the number and extent of anthropogenic fires have dramatically increased, breaking record after record (WWF, 2020), probably caused by recent changes in Brazilian environmental policy. Indeed, the fire events in the Pantanal changed from natural or low human manipulation disturbances to less predictable and intense disturbances (Pivello et al., 2021). Therefore, in such extreme scenarios, the Pantanal fire events can be considered a more acute and unpredictable disturbance than the seasonal floods (chronic disturbances).

Ants are widely used to evaluate the effect of environmental disturbances, and their responses to different disturbance types can vary at different spatiotemporal scales (Andersen, 1999). For example, some previous studies have shown that fire changes the structure of ant communities even after a few years (Maravalhas and Vasconcelos, 2014; Parr et al., 2004). However, rapid community recovery (less than a year) has been observed in other studies in savannas and semiarid regions (Andersen et al., 2014; Parr and Andersen, 2006). In floodplains such as the Brazilian Pantanal, ants of flood zones are constantly subjected to flood disturbances (Dambros et al., 2018). Depending on the elevation of a given location, the flood period may vary from days to months but are generally predictable, occurring seasonally (Junk et al., 2006). While the Pantanal ant communities face two kinds of disturbance, with different impacts and predictability, their response to these joint disturbances is still poorly known.

Although ants may respond differently to fire and flooding in the Pantanal, these responses can be related to time since these disturbances. Most ground-dwelling ant species forage on the trees and surrounding vegetation or change their foraging territories to upper elevations during a flood (Adis, 1997). However, after the plains dry out, these ants return to their usual foraging sites in the soil/ litter strata, now with a simplified structure, which may force them to increase their foraging rates. Indeed, a similar situation happens soon after a fire, and ground-dwelling ant colonies often increase their activity rates (Parr et al., 2007) also due to a simplified habitat and a decrease in the abundance of litter-dwelling arthropod prey (McGlynn et al., 2003). These two factors combined can force the ant colonies to allocate more foragers and forage longer distances (Sanders and Gordon, 2002), and this should be even more intense in the already simplified dry floodplains.

After a longer period, the main effects of the disturbances are not on ant activity but rather on their diversity patterns and community structure (Andersen, 2018). Although the Pantanal ant communities are presumably adapted to seasonal floods and occasional low-intensity fires, the same may not be true to high-intensity and unpredictable human-induced fires. Since chronic disturbances often act as an environmental filter to ant communities, we may expect their effect to be perennial, leading to uniquely adapted ant communities. However, we should expect the ant community to have a more intense response to acute disturbances, at least temporally, such as fires, which can extirpate less adapted species and change overall species coexistence patterns (Rosa et al., 2021). Therefore, if there is a synergistic effect of fire and flood on ant communities, it should be lowered with time, as the community recovers from the acute disturbance.

Here, we assessed the effects of acute and chronic disturbances in Pantanal ant communities, asking whether these disturbances synergistically affect these communities. We expected the fire effect (acute disturbance) to be punctual and to disappear over time, while the flood effect (chronic disturbance) would remain constant, thus: (1) the effects of both (flood and fire) will act synergistically in the short-term post-fire, positively affecting the number of local species and changing ant community composition; (2) in the medium-term post-fire, we still expected a synergistic effect, but with a decreased effect of fire on community richness and composition; (3) in the long-term, the post-fire community would stabilize, and only the effect of flood would be observed on ant community richness and composition.

This study is the first, to our knowledge, to evaluate the relationship between both fire and flood disturbances in ant communities. Contrary to what we expected, we did not find a synergistic effect between fire and flood. Instead, ten days after a fire, we observed a dominant effect of fire disturbance, increasing the number of ant species in recently burned areas and not showing interactions with flood duration. Rather, the combined effect of fire associated with flooding was equal to the effect of fire alone (dominant factor). While the effects of dominant disturbances over many organisms are usually of long-term (e.g., Bruland et al., 1991; Vadeboncoeur, 2010), we found that the dominant effect of acute disturbance (fires) over chronic disturbance (floods) was present in the short and medium-terms, but disappeared with time. This suggests the resilience of the ant communities to this acute disturbance (fire).

We observed a fast increase in species richness immediately after the fires, a common response of many organisms, including ants (Andersen, 2018). This species richness increase soon after fire was accompanied by a marked difference in ant species composition between burned and unburned plots, with ten ant species found exclusively in the former ones. While fire can cause direct and indirect impacts on ant communities (Andersen, 2018; Arruda et al., 2020), the direct effects are often related to the death of colonies (Vasconcelos et al., 2017), which was probably not the case, due to the increase in ant species richness. Meanwhile, indirect effects are more likely to explain this rapid local increase in abundance and richness: ant colonies often increase their foraging activity after a fire due to habitat simplification and a decrease in ground-dwelling arthropod prey (Parr et al., 2007), which can be substantially affected by fires in the short-term (York, 1999). Therefore, the ground-dwelling ants need to allocate more workers to foraging activities after a fire and forage to longer distances (Sanders and Gordon, 2002). Notably, the higher ant richness was positively correlated with the frequency of occurrence, which was also higher in short-term burned areas.

The effect of fire on ant species composition and richness disappeared entirely after four years, indicating that although fire is a major disturbance in floodplains, its effects are temporary. Indeed, low-intensity fires do not seem to have many long-term effects on soil invertebrates, including ants and their prey (Andersen et al., 2005). After the recovery of the vegetation structure, including the grass-layer and the surrounding shrubs and small trees and consequently of soil and litter arthropods,’ the foraging activities of ground-dwelling ants may get back to pre-fire levels. Importantly, the Pantanal is a fire-prone ecosystem, and as such, it has many organisms adapted to fire events, at least to periodical and low-intensity ones (Manrique-Pineda et al., 2021). However, under increased fire frequency in the Pantanal, even fire-adapted organisms may face severe consequences (Pivello et al., 2021). Our results have implications for monitoring Pantanal biodiversity: ant communities can capture the signal of fires regardless of flood exposure, but only in the short and medium-term. Moreover, if the study purpose is not related to the effect of fire on ant communities, our data indicates that the interval of four years after a fire disturbance appears to be sufficient to disconsider its effect.

Contrary to our expectations, we did not detect any effect of the flood period on species richness and composition between the higher or lower elevation areas. This suggests that the species found in flooded areas are not a random subgroup of common species in adjacent non-flooded savannas and forest areas. Instead, the Pantanal wetland ant communities seemingly consist of ants with a high ability to colonize habitats that become seasonally available, including upper areas, like forest patches (Ribas and Schoereder, 2007) and even trees (Soares et al., 2013). Importantly, the Pantanal plains have been subject to flooding for millions of years (Assine and Soares, 2004), and it is known that this environmental filter has shaped its habitat structure and plant species distribution (da Silva et al., 2020). Therefore, ant resistance to floods can be interpreted as a result of their adaptations to a chronic and cyclical disturbance (Orians, 1975), which can act as important species filters (Drapeau et al., 2016), demonstrating that ants can adapt to disturbances (Burns et al., 2020).

Although many studies concluded that disturbances are important for community structuring, few have evaluated how chronic and acute disturbances can interact to structure them at multiple time scales. This gap can be explained by the complexity of experiments with this approach (Kéfi et al., 2019). For example, the unplanned fire events have started in neighboring farms, and were spatially determined in a large spot that hit nearby points. Thus, the spatial correlation at the local scale of the interest factors is common, as we detected in the model with the short-term fire effect. Indeed, as this is a natural experiment (sensuGotelli & Ellison, 2004), techniques to explain the spatial autocorrelation would also capture the fire effect since they are inseparable. Despite these potential limitations, since wetlands such as the Pantanal are facing increasing anthropogenic disturbances, studies on the effects of synergistic disturbances in such endangered habitats are timely. However, wetlands such as the Pantanal are facing increasing anthropogenic disturbances, and studies on the effects of synergistic disturbances in such endangered habitats are timely, despite potential limitations. The observed resilience of the Pantanal ant communities can have important implications since some of their ecological functions, such as nutrient cycling, might also be resilient to fire perturbations. Nevertheless, the studied ant community took some years to recover, and, thus, the increasing frequency of human-induced fire in the Brazilian Pantanal (Pivello et al., 2021) may impose a severe threat to ant diversity and their roles.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.