Fish mortality in the Amazonian drought of 2023: the role of experimental biology in our response to climate change

In 2023, the Amazon basin experienced a compound drought–heatwave event; this record-breaking drought resulted in the loss of 3.33 million hectares of surface water (Espinoza et al., 2024). Between 24 September and 26 October, the surface area of Lake Tefé decreased by 75%, from 379 to 95 km², with large areas less than 0.5 m deep (Fleischmann et al., 2024 preprint). At the Port of Manaus, water levels dropped from a seasonal high of 28.3 m to a record low of 12.7 m, the lowest since measurements began in 1902 (Espinoza et al., 2024). The Amazon also experienced record high air temperatures during this time (Espinoza et al., 2024). This drought, intensified by climate change, caused a mass mortality event, killing large quantities of Amazonian fish and other wildlife (Fig. 1A–D); based on the results of work over recent years determining the critical thermal maximum (CT_max; see Glossary) of many Amazonian fishes (Table 1), we conclude that these species have reached their thermal tipping point (see Glossary).

One of the contributing factors to the record drought in the Amazon was the powerful El Niño phenomenon (see Glossary; Perkins-Kirkpatrick et al., 2024), which exacerbated the unprecedented global temperatures experienced in 2023. However, although the El Niño in the Pacific and the warming of Atlantic waters partly explain the extreme climate conditions experienced, they do not fully account for the crisis. Deforestation in the Amazon, mainly from forest fires, has significantly disrupted rainfall patterns (Reis et al., 2021). This deforestation reduces evapotranspiration (see Glossary) and hampers water condensation, as a result of airborne soot, hindering precipitation. Although the Amazon is naturally fire resistant, human activities and climatic anomalies have increased its susceptibility to fires, which directly affect the ecosystem through runoff (Gomez Isaza et al., 2022).

Faced with the historic extreme drought in the Amazon in 2023, the message is clear: climate change is due not only to natural variation such as the El Niño phenomenon but also to human actions, and the Amazon is already suffering harmful consequences. In fact, rising water temperatures have also been the cause of other mass fish kills around the world. In Australia, millions of fish died in 2023 in the lower Darling Baaka River in the far west of New South Wales, as hot weather in the region exacerbated hypoxia events. Similar events occurred along the Texas Gulf Coast, where the impacts of hypoxia associated with high temperatures resulted in tens of thousands of dead fish being washed up along the shoreline.

In this Perspective, we aim to highlight the urgent need for a deeper understanding of the physiological impacts of extreme weather and climate change on wild species, using the example of the historic Amazon drought of 2023. This Perspective underscores the importance of integrating physiological insights into climate change research and policymaking to protect ecosystems and prevent future environmental crises.

We believe that the compilation of data on the thermal tolerance of Amazonian species – data that have been collected over multiple years – provides a good example of the importance of experimental biology in understanding the responses of wild populations to climate change. During the unprecedented Amazonian drought of 2023, we conducted measurements of water temperature and oxygen levels in Lake Janauacá, situated along the Amazon River in Brazil. On 10 October 2023, remarkably high water temperatures (exceeding 38°C) were recorded. For reference, the surface water of Amazonian lakes typically averages around 29–30°C: from 1995 to 2011, an average temperature of 30.4°C was recorded across 322 samples from 24 different Amazonian floodplain lakes (see Glossary; SO-HYBAM, https://hybam.obs-mip.fr/). In other areas of the Amazon, such as Lake Tefé, temperatures of 40°C were recorded during the extreme drought, even at depths of up to 2 m (Fleischmann et al., 2024 preprint). Such temperatures significantly surpass the CT_max previously established for numerous Amazonian fish species, indicating that the rise in water temperature is responsible for the mortality observed among these animals as they reached their thermal tipping point (Fig. 2). Over the past five decades, the maximum temperature during heat waves in the Amazon has increased by 3–3.5°C (IPCC, 2023); however, the IPCC predicts that the increase in temperature caused by climate change could reach up to 7°C above the average in the next century (IPCC, 2023), which puts the existence of many of the fish species in the Amazon at severe risk (Fig. 2).

It is well known that thermal acclimation (in the laboratory setting) or acclimatization (under natural conditions) to high temperatures can increase thermal tolerance through the remodeling of several physiological and biochemical parameters, including the adjustment of metabolism (Jung et al., 2020; Somero, 2010). Unfortunately, Amazonian fish reside in an environment with low thermal variation and are already living at temperatures that are very close to their upper thermal limits, and they have limited capacity for thermal acclimation or acclimatization (Campos et al., 2019; Lapointe et al., 2018). Despite this, as shown in Table 1, there is some evidence of higher acclimation temperature (which was set based on habitat temperature) increasing CT_max for some Amazonian fish species (e.g. Otocinclus spp., Hemiodus gracilis and Corydoras schwartzi). Acclimatization is observed across the different Amazonian environments; for example, some species in the Amazonian forest stream (see Glossary; ∼26°C) have lower CT_max compared with those found in floodplain lakes (28–31°C; Table 1). Thus, Table 1 shows that the temperature at which the fishes will be severely affected is ∼37–40°C; even the more thermally tolerant species, such as Colossoma macropomum, present mortality at temperatures close to 42°C, regardless of their thermal acclimation regime.

Climate change has a multitude of effects, creating a range of stressors that are further compounded by phenomena such as El Niño or by human activities such as deforestation and forest fires, which increase CO₂ emissions. The predicted increases in CO₂ concentration will lead to decreased water pH, making Amazonian waters more acidic and worsening hypoxia events. Combined with high temperatures, these factors will significantly challenge fish survival. Amazonian lakes already experience daily and seasonal periods of hypoxia, which are exacerbated by rising water temperatures, further reducing oxygen solubility. In parallel, higher temperatures increase a fish's demand for oxygen, as metabolic activity is increased (Pörtner, 2021). Thus, the higher water temperature associated with hypoxic Amazonian waters would make it impossible for a fish's cardiovascular system to efficiently deliver oxygen to its tissues. In addition, an increased metabolic rate is associated with increased oxidative stress (see Glossary) at higher temperatures (Pörtner, 2021), which inhibits an animal's ability to repair injury and causes cellular damage such as protein denaturation, lipid lipoperoxidation and DNA damage (Chowdhury and Saikia, 2020).

As noted above, during the drought–heatwave in the Amazon in 2023, we conducted measurements of dissolved oxygen in Lake Janauacá. The average oxygen saturation across the six measured points was 16.75% (equivalent to 3.8 mg l⁻¹) at an average temperature of 34.6°C. However, we identified some areas with alarmingly low values, reaching as little as 1.59% (0.38 mg l⁻¹) at 31.8°C. The mortality of Amazonian fishes is therefore likely to be due to a combined effect of high temperature and hypoxia. This idea is supported by data on the specific fish species that we recorded among the dead animals on our field expedition. The species that we identified include Cichla monoculus, Hoplias malabaricus, Astronotus ocellatus, Hoplosternum littorale and Satanoperca jurupari. Of those mentioned, we only have CT_max information for A. ocellatus and S. jurupari (Table 1), and these two species present high CT_max values relative to those of other Amazonian species. At Lake Janauacá, temperatures reached 38°C, but they soared to 40°C in certain Amazonian areas during the peak of the drought. It is important to note that CT_max measures the capacity of a species to deal with acute stress, whereas survival at constant temperatures near CT_max is highly improbable. The increased energy expenditure required for the physiological responses needed to compensate for thermal stress is unlikely to be viable over prolonged periods. Furthermore, the mortality of such species can be attributed to a combination of factors associated with the increased temperature. Given the link of the Amazonian fish mass mortality event with both high temperatures and hypoxia, we note that this event may support the theory of ‘oxygen- and capacity-limited thermal tolerance’ (OCLTT), which proposes that thermal limits are related to the inability of the cardiorespiratory system to efficiently deliver oxygen to the tissues.

The simultaneous presence of hypoxia and elevated temperatures not only increases oxygen demands in fish but also reduces their ability to acclimate to higher temperatures. This poses a significant concern for Amazonian fish, which already face limitations in acclimation/acclimatization capacity. Climate change thus has the potential to worsen existing environmental challenges in the Amazon basin, potentially impacting the distribution, abundance and survival of freshwater fish, depending on their adaptability to these changing conditions.

Work performed in the Amazon basin highlights the urgent need to understand and address the combined impacts of climate change. By examining how high temperatures and hypoxia affect fish populations during the extreme drought, we reveal the complex challenges facing freshwater ecosystems. This underscores the importance of a comprehensive approach to ecophysiological research, considering the synergistic effects of multiple stressors. Collaboration among scientists, policymakers and stakeholders is crucial.

This Perspective emphasizes the pivotal role of experimental biology in elucidating the responses of species to climate change, with the 2023 Amazonian drought serving as a prime example. Previous studies on Amazonian fish have demonstrated their limited thermal plasticity and proximity to thermal thresholds, making even slight temperature increases potentially lethal (Campos et al., 2019). Future research should prioritize investigating thermal tolerance and adaptive capacity across ecosystems to anticipate and mitigate the adverse effects of climate warming on biodiversity. This will require interdisciplinary collaboration and innovative methodologies. Understanding species’ functioning is vital for predicting their vulnerability to climate change. Utilizing physiological metrics such as CT_max can inform effective public policies and conservation efforts, as emphasized by Madliger et al. (2021) and Cooke et al. (2021). Further research is required on organisms' complex responses to interactive biotic factors, integrating physiology with conservation (Franklin and Hoppeler, 2021). As climate change reshapes ecosystems globally, it is imperative to prioritize resilience-building strategies that consider the intricate interactions between different factors. Integrating these research findings into conservation frameworks can guide the development of adaptive management and policies to mitigate the impact of climate change on biodiversity and ecosystem health.

Experimental biology has revealed the dire consequences of climate change on Amazonian fish species during the 2023 drought, underscoring the urgent need for comprehensive studies to understand and mitigate future events. Information on species' thermal limits and acclimatization capabilities, gained from experimental biology research in the Amazon, allows us to extrapolate laboratory findings to natural environments and assess fish mortality from the 2023 drought. This knowledge informs targeted conservation strategies, such as the creation of refuges in cooler areas. It also emphasizes the need for global climate action to mitigate rising temperatures and species mortality.

Local leadership in research, supported by global collaboration, is crucial for addressing climate change impacts worldwide. Furthermore, it is of the utmost importance to provide equitable support for researchers from Global South countries, as they are often the most adversely affected by climate change. International collaborations can enhance understanding and inform evidence-based decision making, thereby bolstering resilience against climate change. By using concrete examples to raise public awareness, encouraging community engagement in conservation and advocating for stronger environmental policies, we can address the immediate threats of climate change and protect biodiversity in the Amazon and elsewhere.

The authors would like to thank João Henrique Alliprandini for his help with ideas for creating the graph in Fig. 2, and Thiago Luís Alberto da Silva do Nascimento for his support in field expedition.