El Niño Southern Oscillation (ENSO) events are associated with abnormal surface water warming that occurs at irregular intervals in the tropical Pacific and can affect global climate (Marengo et al., 2012). Severe ENSO seasons affect not only the Pacific Ocean but the inland waters of the entire planet (Ward et al., 2014). In the northern region of Brazil, El Niño episodes cause drastic changes in rainfall patterns and consequently promote more extreme droughts than normal, while La Niña episodes have the opposite effect (Reboita et al., 2021). In Amazonian river systems, ENSO effects are reflected by changes in flow conditions, limnological variables, lateral connectivity, and seasonal biogeochemical cycles (Towner et al., 2021). During El Niño, the absence of rain causes, among others, severe droughts, decreased lateral connectivity, intense forest fires, and reduced floodplain areas. Meanwhile, the La Niña seasons are marked by high hydrological and flow levels, large amounts of rainfall, increased biological productivity, and a diversity of habitats suitable for the development of associated organisms (Towner et al., 2021). Additionally, La Niña seasons are related to abundant fishing harvests in the Amazon (Isaac et al., 2016).

Although the Amazon ichthyofauna is very adapted to the intense hydrological dynamics that naturally occur in this system, due to the periodicity of the flood pulse, studies indicate that extreme climatic events can cause significant and lasting changes in the taxonomic composition and functional diversity of the fish community in floodplain environments (Freitas et al., 2013; Röpke et al., 2017). The way the ichthyofauna responds to the existing environmental conditions during anomalous hydrological events is related to the ecological characteristics of each species and that often allow them to support or not such conditions (Chessman, 2013; Chea et al., 2020). The climatic anomalies associated with ENSO impact several population aspects of fish, especially reproductive ones. For example, prolonged droughts and mild floods can cause high mortality of larvae and juveniles, lower proportions of females with mature gonads, and decreased size at first maturity (Castello et al., 2019; Röpke et al., 2022). On the other hand, some species of periodic and migratory fish benefit from larger nursery areas, feeding areas, and the connectivity between the main channel and the floodplains during La Niña events when the flood is intense (Oliveira et al., 2014; Camacho Guerreiro et al., 2021; Cataldo et al., 2022).

Although studies indicate the existence of the effect of La Niña and El Niño on the reproductive success of ichthyofauna, few investigations have been carried out on this topic, especially in the Amazon. Empirical information from the perspective of the ichthyoplankton community stands out as a fundamental tool to assess the effect of changes in environmental gradients on fish reproduction (Cajado et al., 2020; Tyler et al., 2021). Ichthyoplankton dynamics reflect the status of commercially and ecologically important fish populations and provide predictive answers about potential future recruitment (Zacardi et al., 2017a; Humphries et al., 2020).

Understanding the response of the fish community to climatic anomalies through a functional face applied to ichthyoplankton becomes fundamental because it makes it possible to identify a more predictable response of biological communities to environmental changes (Arantes et al., 2019a; Gogola et al., 2021). Information of this nature would contribute to the understanding of the effect of climate change on the reproductive aspects of Neotropical fish. This fact is aggravated because, in recent decades, events such as ENSO have become increasingly intense and frequent in the tropics due to increasing changes in the global climate (Espinoza et al., 2019); and thus, studies of their effects on ichthyoplankton communities can help in decision-making for effective management measures for the conservation of fisheries resources.

In this study, we evaluated the possible influences of ENSO on the taxonomic and functional structure of the fish larvae community in the surroundings of a Lower Amazon River floodplain. Environmental variables with potential influence on the structure of ichthyoplankton assemblages were also included in our analyses. We collected ichthyoplankton in three hydrological periods: neutral, La Niña, and El Niño during the rising of river water, the reproductive phase of most fish species in the Lower Amazon (Zacardi et al., 2017b), to test the hypothesis that anomalous climatic events significantly affect fish reproduction, with changes in demographic aspects and in the taxonomic and functional structure of fish larvae assemblages. We expect that the response of species to climate change will be influenced by their functional traits. For example, pelagic fish larvae must be associated with different periods than benthic species, because the former may be more sensitive to changes in water masses than the latter, which have more geographically fixed, near-bottom habitat requirements (Duarte et al., 2020, 2022; Silva et al., 2020). In addition, periodic species, as well as medium-distance migrators, should be more abundant during La Niña, due to favorable conditions for the reproduction of these fish, such as greater amounts of rainfall and river flow (Zacardi et al., 2017b; Humphries et al., 2020). Finally, we believe that commercially exploited species will have their reproduction more affected, given that previous studies have observed the effect of climatic events on the reproductive parameters of these fish (Röpke et al., 2022).

Environmental variables differed between the periods studied. The first two axes of PCA explained approximately 55% of the data variability. The La Niña period presented higher values of electrical conductivity, pH, and rainfall that were positively associated with PC1, while El Niño presented higher values of temperature, ONI, dissolved oxygen, and flow, variables associated more with PC2, except for flow that was negatively more associated with PC1 (Figure 2; Table 1). The neutral period recorded a gradient associated with the intermediate conditions environmental between the two periods of ENSO.

A total of 50,484 fish larvae classified in nine orders, 27 families, 40 genera and 55 species were captured (Supplementary Material S1). The composition and abundance of larval assemblages differed significantly between the neutral (2013 and 2014), La Niña (2018) and El Niño (2019) periods (LR = 763.5, p-value = 0.001) (Figure 3). The assemblage differences between each period were driven by the significant contributions of 20 species (Figure 4). The mean density of larvae captured in the neutral period (123.05 larvae 10 m⁻³) was 13 times greater than the La Niña period (17.88 larvae.10 m⁻³) and about six times greater than the El Niño period (39.88 larvae.10 m⁻³).

ONI and local environmental variables explained 9.66% of the change in species composition and had, except temperature and dissolved oxygen, significant effect on fish larval assemblages (RDA: permutation tests, df = 7, F = 3.55, p-value < 0.001). The first two axes of the RDA explained 13.44% of the restricted variation. The taxonomic structure of the larval assemblage was strongly associated with the environmental conditions arranged along the temporal gradient (Table 2; Figure 5). The larvae of the Characiformes Mylossoma aureum, M. albiscopum, Schizodon fasciatus, Leporinus sp., Anodus elongatus, Siluriformes Pseudoplatystoma punctifer, Sorubim lima, and Clupeiformes Pellona flavipinnis, P. castelnaeana and Engraulidae gen. sp. were more representative in the neutral period and showed a sharp decline in their abundances in subsequent periods (LR: p-value < 0.05) (Figure 4). Interestingly, all these species, except Engraulidae gen. sp., are highly commercially exploited in the Lower Amazon region. The curimatid Psectrogaster amazonica we associated with the La Niña period and were positively associated with high gradients of electrical conductivity (LR = 42.78, p-value = 0.001) and pH (LR = 12.43, p-value = 0.04). While Auchenipteridae gen. sp., Pimelodus blochii, and Hemiodus microlepis increased in abundance during the El Niño period, a fact evidenced by the association of these species with high ONI (Figure 5). Furthermore, the abundance of these species decreased with higher pH values (LR: p-value < 005).

The functional structures of the assemblages also changed between periods (LR = 739.4, p-value = 0.001) and were strongly modulated by ONI and local environmental variables (RDA: permutation tests, df = 7, F = 4.53, p-value < 0.001) that explain 11.61% of the change in the functional structure of the larval assemblage. Larvae of fish large periodicals, herbivores, piscivores, epibenthic maneuverable, and nektonic burst swimmers were more abundant during the neutral epoch and were negatively associated with electrical conductivity and pH and positively to the flow (p-value < 0.05) (Table 3). Larvae of fish migratory medium-distance, with a life history strategy periodic of small-sized, and maneuverable nektonic predominated during the La Niña season (2018), in addition, they were positively related to the high gradients of pH, electrical conductivity, and negatively to the flow. While larvae of species with intermediate life history strategy, sedentary, short-distance migratory, planktivores, omnivores as well as benthic slow and benthic fast species, were associated with El Niño (2019), a fact observed by the positive correlations of these groups with high ONI, in addition to being positively related to higher values of flow and temperature and negatively to rainfall, electrical conductivity, and pH (p-value < 0.05) (Figures 6A,B). Larvae of fish opportunistic life-history, as well as long-distance migrators, carnivores, and Surface dwellers, were not affected by the periods. However, they were negatively influenced by pH and electrical conductivity. Only long-distance migrants and Surface dwellers were not influenced by the variables included in our models.

The patterns of association of larval community to hydrological periods are dependent on the ecological strategies of each species and show that climatic events modulate the structure of larval fish assemblages, corroborating our hypothesis. These findings contribute to an understanding of interactions between anomalous climatic events and fish reproduction in the Amazon River. Larvae of large periodic fish, piscivores, herbivores, maneuverable epibenthic, and nektonic burst swimmers predominated in the neutral period. On other hand, larvae of small periodicals and medium-distance migrators were more abundant in the La Niña period. In addition, larvae of fish with basal feeding habits (planktivores), generalists (omnivores), intermediate strategists, short-distance, and sedentary migrators, as well as benthic and commercially unexploited species, dominated assemblages during El Niño.

The environmental conditions were characteristic of each period, but the flow and precipitation values diverged from the proposed models for Amazon region throughout the seasonal cycle (Towner et al., 2021). Higher flows were observed during El Niño despite lower amounts of rainfall. Recent studies have documented abnormal precipitation in Andean regions caused by Pacific Decadal Oscillation (PDO), and ENSO in positive phases (Rodríguez-Morata et al., 2019; Mohammadi et al., 2020). This may favor higher temperatures in this period and, consequently, rapid melting of glaciers that contribute as one of the sources of water for the Amazon River, as observed by Rocha et al., 2019. It is known that the Andean thaw acts as a regulator of the hydrological regime and strongly affects geomorphology, biochemistry, and ecology of headwater tributaries of the Amazon basin (Filizola and Guyot, 2011; Vauchel et al., 2017), exporting a source of sediments, organic matter, and nutrients to the lower sectors of this basin.

The higher values of pH and electrical conductivity during La Niña possibly associated with high rainfall recorded during this period. The rainfall regime is a main regulatory agent of water courses, being responsible for the distribution of nutrients and ions in the water column, affecting concentrations of physicochemical variables in rivers (Esteves, 1998; Zuliani et al., 2016). According to Carvalho et al. (2000), during increased precipitation, pH tends to rise and approach neutrality, as there is a greater dilution of dissolved compounds and faster flow. This is caused by the increase in the volume of water which causes the acidity of the water to decrease. However, rainfall is responsible for leaching the soil from adjacent regions, influencing the periodic supply of suspended solids, particulate inorganic matter, and natural calcium carbonate in the region, and increasing turbidity (Silva et al., 2008). Thus, the electrical conductivity that depends on the ionic composition of water bodies is also influenced by the volume of local rainfall.

Our results showed that in the neutral period the mean density of larvae was much higher than in the La Niña and El Niño periods. This high variability in abundance between different periods has also been recorded in other river systems (Bednarski et al., 2008; Marschall et al., 2021). Changes in the flow and amount of rainfall have been suggested as plausible explanations for the variation in fish reproductive success in rivers with floodplain areas (King et al., 2009; Andrews et al., 2020; Stoffers et al., 2022). In fact, most species associated with the neutral period showed a positive relationship to flow, pH, and electrical conductivity, but precipitation was less important. In Amazonian rivers, increased flow, which is correlated at river level, is the main mechanism for transporting eggs and larvae to floodplain areas (Oliveira et al., 2020; Zacardi et al., 2020). It is noteworthy that the early stages of fish are critical due to their high susceptibility to predation and the need for readily available food resources for growth (Grenouillet and Pont, 2001; Montenegro et al., 2020). Therefore, the floodplain is considered an excellent nursery and natural breeding ground during the initial cycle of fish development (Bittencourt et al., 2020; Zacardi et al., 2020).

The studies by Baumgartner et al. (2008) and Tondato et al. (2010) also recorded the association of pH and electrical conductivity with the abundance of fish larvae, however, the effect of these variables on fish reproduction is not known for sure, but somehow, they induce migration and spawning for some species. Additionally, the decrease in larvae of fish of commercial interest after the neutral period may indicate some connection with fishing carried out in the region. Zacardi et al. (2018) verified the relationship between the production of landed fish and the abundance of larvae in the lower stretch of the Amazon River and observed that the fishing effort undertaken on adults compromises the spawning stock and, consequently, the annual recruitment of resources. The reproductive parameters of commercially exploited species are more susceptible to the effects of climatic events (Röpke et al., 2022), suggesting that environmental factors and fishing intensity can act simultaneously for the failure or success of fish recruitment in each period. Therefore, relatively minor changes in the reproductive success of large periodic can lead to poor recruitment for at least 2 years (Bayley et al., 2018; Castello et al., 2019; Cataldo et al., 2022).

The inferred effects of climatic events on the biodiversity of fish larvae indicate that the environmental conditions associated with ENSOs act as a filter in the functional structure of local assemblages, as observed in studies involving adult fish in the Amazon (Freitas et al., 2013; Röpke et al., 2017). Several functional groups had reproduction strongly associated with periods of high flux values, as they exploit favorable conditions for the transport of larvae to locations conducive to development. The large-size periodics decreased in abundance during ENSOs. These fish are more exploited by fisheries, in addition, they have a slow life cycle, being more sensitive to environmental changes (e.g., decreased flow) and anthropogenic pressures (Castello et al., 2019; Camacho Guerreiro et al., 2021). Larvae of many fish epibenthic maneuverable, herbivorous, and piscivorous benefit from increased flows as they are carried by rising water levels into recent floodplains to explore structurally complex habitats. For example, herbivore larvae (e.g., Mylossoma spp.) tend to colonize aquatic vegetation in search of refuge and food (Bittencourt et al., 2020; Oliveira et al., 2020), and piscivorous larvae maximize their survival by capturing abundant prey in the various habitats and micro-habitat of the floodplain (Tondato et al., 2010; Nunn et al., 2012; Grzybkowska et al., 2018).

On the other hand, although the lower flows during La Niña may indicate some disadvantages in dispersal mechanisms, the offspring of small periodic fish can take advantage of the low abundances of predatory fish larvae, such as piscivores and large periodic, to maximize recruitment. This strategy guarantees good ecological conditions and advantages for the survival of small periodic since predation is one of the main mortality factors during the initial ontogeny of fish (Nunn et al., 2012; Cajado et al., 2018; Humphries et al., 2020).

The El Niño period, in turn, was less favorable to the reproduction of some previously abundant functional groups (e.g., periodicals, medium-distance migrators, herbivores, and maneuverable epibenthic), suggesting that the combination of environmental conditions favorable to the reproduction of most of these groups did not occur effectively. This result corroborates the view that, regardless of anomalous hydrological periods, there is scaling at the time of spawning season between species (Shuai et al., 2016; Barthem et al., 2017; Tyler et al., 2021) and functional similarity reflects the preference between reproductive periods. For example, the associations of slow benthic and fast benthic fish larvae, as well as sedentary ones with the high fluxes in El Niño corroborate previous studies that demonstrate that species of these functional groups require minimum hydrological levels to trigger their reproduction (for example, greater than 5 m) (Cañas and Waylen, 2012; Hermann et al., 2021). This allows the larvae to enter the floodplain areas and explore the spatial patterns and nature of resources essential for successful recruitment, as reported by Duarte et al. (2020, 2022) and Zacardi and Ponte (2021). Thus, the period and extent of the flooding over the river plain significantly contribute to the generation and maintenance of the biodiversity of the Amazon ichthyofauna (Röpke et al., 2017; Frederico et al., 2021).

These examples further indicate that changes in environmental conditions associated with ENSOs favor some functional groups but are detrimental to others. However, a deeper understanding of how climatic events influence the structure of fish descendants can be obtained from the expansion of the sampling effort in spatial and temporal scales, inclusion of interactions between factors such as seasonal periods, water level elevation intervals, river, prey availability, lateral connectivity, and forest cover.

The variation in the taxonomic and functional structure of the fish larvae community between periods highlights the potential for extreme weather events as a factor affecting fish reproduction in the Amazon River. Despite this evidence, further studies are needed to better understand the processes that shape the reproductive dynamics of fish under anomalous hydrological conditions in the Amazon basin. Although we can only assume, the interaction between hydrological variations and factors such as changes in local environmental variables, ecological relationships, and fisheries seem to enable the success of certain functional groups in each period. The findings of this study can serve as a baseline to assist the implementation of protective measures on reproductive fish stocks, for example, in times of ENSO under atypical flow conditions and changes in environmental characteristics, where the intensity of reproduction and the biological recruitment is predictable, and thus adjust the duration of fish reproduction protection periods. An alternative widely used in the region as an effective action for the conservation of species and the sustainable use of fisheries resources.