Abstract
Although climate change may facilitate the spread of invasive alien species (IAS), research assessing this link remains limited. Recognizing suitable habitats where IAS have been introduced is crucial for biodiversity conservation and ecosystem management. Here, we integrated online, museum, and laboratory occurrence databases with local ecological knowledge (LEK) on IAS fishes from semi-structured interviews and georeferenced social network posts from traditional and recreational fishers to model the habitat-suitability of three voracious IAS Cichlids introduced in the Brazilian part of the Upper Paraguay River Basin (the Pantanal wetland and its tributaries). Our goal was to locate areas (regions, sub-basins, and reservoirs) most at risk from the spread of these IAS fishes in the basin. The findings reveal extensive suitable habitats for these IAS fishes throughout the study basin, currently covering half of the Pantanal wetland and up to 90% of the upstream sub-basins. Under future climate scenarios, these suitable habitats are projected to expand further, encompassing almost the entire Pantanal floodplain. These results highlight a potential IAS Cichlid range expansion in the Pantanal floodplains in the upcoming decades, correlating with common climate change forecasts. We also emphasize the value of integrating ecological niche models (ENMs) along with Citizen Science data to identify high-risk areas during early invasion stages, inform preventive strategies, and support conservation efforts to mitigate the impacts of IAS on native biodiversity.
Similar content being viewed by others
Introduction
Climate change and biological invasions pose significant threats to global biodiversity and ecosystem services1,2,3. Human activities have introduced over 37,000 invasive alien species (IAS) worldwide, contributing to 60% of documented global plant and animal extinctions4,5. In addition, IAS impact food security and human health through the spread of pests and diseases, often exacerbated by changes in climate patterns, resulting in an estimated global economic cost of USD 423 billion annually6,7. The specific role of climate change as a key driver of IAS proliferation is now widely recognized, as rising temperatures and changes in precipitation have altered ecosystems and created novel, suitable habitats for establishing, sustaining, and spreading IAS populations8,9. This pattern is expected to intensify, with important short-term impacts projected by 205010,11.
Freshwaters are among the most vulnerable ecosystems to climate change and IAS, given their limited extent, strong connectivity dependence, restricted species dispersal, and direct dependence on climate stability12,13. Furthermore, freshwater species face the highest extinction rates compared to their terrestrial and marine counterparts, with a quarter of freshwater fauna at significant risk of extinction14,15. In the Americas, particularly in Latin America, freshwater habitats face severe hydrological alterations, declining water quality, and ecosystem degradation, along with the impacts of IAS16,17. Fish species, the most commonly documented IAS in Latin American rivers, lakes, and wetlands, have rapidly spread through aquaculture, posing a major concern in freshwater ecosystems18,19. Several IAS fish species support profitable fisheries despite ecological risks, particularly in the Neotropical region20. However, in developing countries like Brazil, managing IAS fish is hindered by data gaps, socio-economic interests, and limited state intervention21.
The Upper Paraguay River Basin (UPRB) is situated at South America’s geodetic center and spans Brazil, Bolivia, and Paraguay. It encompasses the extensive Pantanal wetland, characterized by intricate floodplains nourished by upstream tributaries of the Paraguay River, supporting one of the richest freshwater fish faunas (≈ 300 spp.)22,23. The UPRB is particularly affected by numerous hydropower plants and reservoirs, with the development of more anthropogenic developments currently planned24. Over the past seventy years, the construction of dams has disrupted freshwater ecosystems by altering sediment flow and water chemistry composition, ultimately obstructing fish migration routes and altering fish assemblages25. Furthermore, hydroelectric reservoirs have provided favorable conditions for the proliferation of IAS fishes26,27. The ecological consequences associated with dam construction, including the proliferation of IAS in reservoirs, must be considered and understood28.
Cichlids are the best example of potential IAS fish establishing themselves in the UPRB. Since 1982, the Blue Peacock Bass, Cichla piquiti Kullander & Ferreira 2006, a Cichlid introduced from the Araguaia-Tocantins basin, has been reported in the Brazilian stretch of the UPRB (hereinafter Br-UPRB)29,30. This species has progressively expanded its invasion range through the Pantanal wetland and some sub-basins like São Lourenço-Vermelho, Miranda, and Apa-Perdido31. Moreover, the finding the Yellow Peacock Bass, Cichla kelberi Kullander & Ferreira 2006, a second Araguaia-Tocantins Cichlid in lotic ecosystems in the basin, has raised concerns regarding these two IAS within the basin30. The Nile Tilapia, wich is another Cichlid with great invasive potential, has also been reported in the basin32,33. While Oreochromis niloticus Linnaeus 1758, is cataloged as an IAS in several tropical regions and considered highly detrimental to native fauna;, it has nevertheless been extensively cultured in local reservoirs of the Br-UPRB along Mato Grosso and Mato Grosso do Sul States recently32,34,35. Frequent escapes from fish farms have increased the presence of this IAS in river catches by local artisanal and recreational fishers (C.M-G Pers. Obs). While the short-term effects of IAS Cichlids proliferation are well documented, such as their role as aggressive predators and competitors that disrupt trophic webs, the long-term ecological impacts still remain uncertain28,36,37,38.
In recent years, Ecological Niche Models (ENMs), a set of correlative techniques analysis known as has experienced a significant increase in popularity to identify the potential extent of IAS invasive ranges4,39. The rising implementation of this approach is partly due to the ability to integrate IAS occurrences with the recently available high-quality environmental data, also informing the magnitude of climate change impacts8,40. ENMs are based on the ecological niche concept, which posits that a species will have a greater preference for a location where the environmental conditions necessary for its survival are maximized41. In addition, ENMs have proven effective in evaluating fish species distribution, contributing to native conservation and non-native population control42.
The ENMs assess the abiotic component (A) of the BAM framework (geographic distributions of species determined by: Biotic, Abiotic, and Mobility features) by analyzing the correlation between species occurrences and environmental variables43,44. While ENMs do not directly incorporate biotic interactions (B) or dispersal limitations (M), these models provide valuable insights into the ecological preferences constraining species distributions41, i.e., the fundamental niche as the range of abiotic conditions where a species can potentially sustain viable populations43. On the other hand, the realized niche is further constrained by B and M45. These ENM techniques allow us to study past, present, and future species distributions, predicting the IAS invasion spread patterns by identifying the most suitable areas for the species in the early stages of invasion, thus allowing us to prioritize efforts in critical regions46,47,48,49.
The use of geographic information systems (GIS) has been improving the interpretation of ENM outputs by considering biogeographical and ecological constraints, including dispersal barriers and habitat fragmentation (affecting M), the effect of IAS (affecting B), and climate change (affecting A)50,51. This integration is particularly valuable for getting closer to the actual species distribution ranges, especially in freshwater ecosystems that function as ecological networks or island-like habitats42,52,53. When ENMs are used to model the distribution of IAS, native range data is used to identify species’ environmental tolerances46,54. In contrast, invaded range occurrences may reflect ongoing colonization processes, founder effects, or incomplete habitat exploration46. This approach helps to understand where invasive species might expand based on their known preferred ecological conditions, facilitating more reliable identification of suitable habitats in invaded regions, thus supporting invasion risk assessments and management actions in vulnerable ecosystems55,56.
The Shared Socio-economic Pathways (SSPs): SSP1 1.9–2.6, SSP2 4.5, SSP3 7.0, and SSP5 8.5, consist of five narratives outlining potential changes in demographics, economics, technology, society, governance, and the environment over the ongoing century57. These pathways serve as a tool for comparing ENMs in scenarios of temperature increase, ranging from 1.5 °C in the near future (2040–2070), to 3.3 –5.7 °C in the distant future (2081–2100), assuming the greenhouse gas emissions and inequality scenarios persist58. Programs such as Citizen Science bride gaps between scientists and local ecological knowledge (LEK) of IAS, providing a cost-effective approach that engages community members in data collection and assessment of IAS impacts on fishing communities59,60,61. The Citizen Science data has been incipient, nevertheless, effectively implemented in some cases of ENMs and IAS management62,63,64. Estimating the potential distribution of the IAS fishes in the Br-UPRB is essential to support management actions, especially considering the predicted and ongoing climate change that threatens freshwater fish native species23.
Using an ensemble of ENMs, we modeled the current and future (2050 - SSP2 4.5) suitable habitats for O. niloticus, C. kelberi, and C. piquiti across their native and introduced (Br-UPRB) ranges. The models integrated hydroclimatic and topographic predictors with IAS occurrences data from online, museum, and laboratory occurrence databases, supplemented by semi-structured interviews and georeferenced social network post, specifically, Instagram posts from traditional and recreational fishers. This study aimed to (i) identify the most suitable habitats for IAS Cichlids within Br-UPRB reservoirs and key regions, including wetland complexes in low-elevation areas (Pantanal), tributary sub-basins on the Brazilian plateau, and (ii) evaluate whether these suitable areas may expand under the 2050 - SSP2 4.5 climate scenario. Given the Nile and Araguaia-Tocantins basins’ origin of the IAS Cichlids introduced in the Br-UPRB, we expect that climate change could expand the climatically suitable areas for these species. Specifically, we hypothesize that combining Pantanal wetland and reservoir conditions (e.g., stable water levels, reduced flow velocity, clear waters) and projected rising temperatures may increase favorable environments for these IAS to spread along the Br-UPRB. This hypothesis is supported by evidence that lentic ecosystems and reservoirs often facilitate the dispersal of Cichlid invasive species65,66 and that rising temperatures may increase the range of neotropical IAS fishes8.
Results
Selection of environmental variables and their contribution to ENMs
The climatic predictor variables for current and future environmental conditions were divided into two sets: (1) for Br-UPRB + Araguaia - Tocantins basins (for modeling the two Cichla spp. habitat suitability) and (2) for Br-UPRB + Nile basins (for modeling O. niloticus habitat suitability). We then added topographic predictors (elevation, flow, and slope), considering the Variance Inflation Factor (VIF) value < 5, to minimize multicollinearity and improve interpretability of the models (Tables S1-S6). In the ensemble of ENMs, precipitation seasonality was the most important single variable for both Cichla species in current and future 2050 - SSP2 4.5 scenarios. For C. piquiti, precipitation seasonality contributed 30.2% in current conditions and increased to 37.8% in future scenarios. For C. kelberi, precipitation and temperature seasonality accounted for 43.6% and 46.3% of variable importance in current and future scenarios, respectively. On the other hand, precipitation of the warmest quarter and mean temperature of the wettest quarter were the most important variables for O. niloticus, contributing 28.9% in current conditions and increasing substantially to 44.3% in the future 2050 - SSP2 4.5 scenarios (Figures S1-S11).
Calibration of the models, validation, and forecasting performance
The individual models selected to build the ENMs ensemble showed high reliability, with Sensitivity and Specificity consistently above 95%, supporting the predictive capacity to identify suitable habitats (Table 1). The evaluation of True Skill Statistic (TSS) and Area-Under-the-Curve (AUC) metrics also confirmed the good performance of the models selected, with a mean calibration AUC ≥ 0.98, validation AUC ≥ 0.93, calibration TSS ≥ 0.91, and validation TSS ≥ 0.78 for all IAS and climatic scenarios (Table 1 and Figure S14). MaxEnt and RF demonstrated the highest performance values among the individual algorithms across all species and scenarios. (Table S7). The findings of this study suggest that the models are appropriate for the projection of current and future (2050) changes in habitats available for the IAS Cichlids in the BR-UPRB.
IAS suitable habitats in the Br-UPRB
The ENMs ensemble confirm this high proportion of current suitable habitats for all IAS Cichlids in the Br-UPRB. Under current environmental conditions, IAS Cichlids in the Br-UPRB have access to approximately 29,483 km² of suitable lotic and lentic freshwater ecosystems, covering half of the basin (51.9%) (Table 2). C. piquiti is currently the IAS Cichlid introduced in the basin with the most suitable habitats available, and for its future expansion in the 2050 SSP2 4.5 scenario. Collectively, the IAS Cichlids may expand their suitable habitats over 12,713 km2 in the future scenario modeled. These overlapped projections of O. niloticus, C. piquiti, and C. kelberi indicate that the habitats available for these IAS can cover two-thirds of the basin (74.3%) (Table 2).
According to the data, more than half of the Pantanal currently provides suitable habitats for IAS Cichlids (58.10%). At the same time, almost a third of the tributary sub-basins are suitable for these IAS (31.48%). However, future conditions predic that suitable areas for IAS Cichlids may increase to 90.15% of the available freshwater ecosystems in Pantanal and 41.79% of the tributary sub-basins (Table 3). These increases reflect shifts in the suitability of habitat, i.e., the areas that may no longer be suitable, those that may remain stable, and those that may become suitable for these IAS in the future (Table 2; Fig. 1).
By 2050, predicted climate change may influence the availability of suitable habitats for the three IAS differently. O. niloticus is predicted to maintain its suitable habitats relatively stable with moderate expansion in the future (IAS Range Change of 14.7%), showing the most conservative response among the three IAS (Table 2; Fig. 1a). On the other hand, C. kelberi exhibits a more pronounced rate of change (IAS Range Change of 56.7%), with suitable habitats expanding primarily around the sub-basin of introduction to the northwest of the basin (Jaurú-Sepotuba) and extending to the southeast of the Pantanal (Tables 2 and 3; Fig. 1b). Finally, C. piquiti might experience the highest rate of change (IAS Range Change of 248.1%), showing potential expansion from current optimal environments near the middle course of the Paraguay River (main river in Br-UPRB) to the whole flood plain and São Lourenço-Vermelho and sub-basins further south (Tables 2 and 3; Fig. 1c).
Consistent with the general basin-wide trend, the results reflect that the suitability of the habitats may increase in the Pantanal wetland (lowlands) and its tributary sub-basins in the Brazilian plateau (highlands). O. niloticus shows strong suitability in the southern Plateau sub-basins, particularly in Aquidauana-Negro and Apa-Perdido, coinciding with their initial introduction areas (Table 3). While C. kelberi could expand its presence in the Jaurú-Sepotuba sub-basin, where it was originally introduced. C. piquiti exhibits the most extensive suitable habitats in the Pantanal wetland, maintaining dominance in multiple sub-basins, including Miranda, and expanding significantly across the region in future modeled scenarios (Table 3). Although suitable habitats for C. piquiti are trending upwards, some sub-basins may experience a decrease in these habitats (e.g., Taquari-Coxim) or maintain limited suitable habitats (< 10% of the sub-basin, e.g., Cuiabá) compared to the remaining sub-basins. This pattern is explained by the expected turnover of suitable habitats, expressed by the Range Size Change (Table 2; Fig. 1c).
IAS suitable habitats in Br-UPRB reservoirs
The IAS fish species introduced in the Br-UPRB currently exhibit limited suitable habitats in the basin’s reservoirs (15.52% overall), with greater suitability available in large reservoirs. Small reservoirs (1–3 km²) show the lowest current suitability at 10.34%, while medium-sized reservoirs (3–13 km²) present moderate suitability at 18.18%, and large reservoirs (> 13 km²) demonstrate the highest current suitability at 28.57%. Nevertheless, according to our projections for 2050, the extent of suitable habitats for these species is expected to more than double within all three categories of these environments. Most notably, large reservoirs are projected to reach nearly 60% suitability (57.14%), representing a two-fold increase from current conditions. Overall, the total suitable habitat in reservoirs is projected to increase from 15.52% to 34.48%, indicating that more than one-third of all reservoirs in the Br-UPRB may become suitable for IAS establishment by 2050 (Table 4).
Projected changes in habitat-suitability for: (a) Oreochromis niloticus (b) Cichla piquiti, and (c) Cichla kelberi, in the study area (Br-UPRB) between Current and Future 2050 - SSP2 4.5 scenarios, based on ENMs ensemble: Generalized Linear Models (GLM), Random Forest (RF), and Maximum Entropy (MaxEnt): Light blue: areas to be lost, Dark blue: areas to remain unsuitable, Light red: areas to may gain suitability, Dark red: areas to remain suitable, Orange dots: IAS Occurrences. The habitat-suitability maps were created using ArcGIS Pro 2.8: https://www.esri.com/en-us/arcgis/products/arcgis-pro/overview.
Discussion
The Pantanal wetland and its tributary sub-basins exhibit a large extension of suitable habitats that are optimal for the future proliferation of introduced IAS Cichlids, even under one of the most optimistic future climate scenarios (i.e., 2050 - SSP2 4.5)67. Conversely, there is a substantial loss of suitable habitats for native freshwater fish species in the Br-UPRB under climate change23. These opposing trends highlight a critical concern: while the expected environmental conditions may facilitate the thriving of IAS Cichlids, the native fish species may face a significant loss of diversity15,23,68. The potential proliferation of these introduced species assessed in our study may likely intensify the climate change challenges for native species, disrupt biological communities, alter ecosystem structure and function, and lead to important socio-economic consequences35.
The importance of climate seasonality in our ENMs is consistent with previous observations that seasonal temperature and precipitation patterns influence the potential spread of IAS fishes69,70. This antecedent could also be related to the fact that the colonization success of Cichlid species is greatly influenced by abiotic factors, their plasticity, and their high adaptive capacity71,72. Unexpected changes in the climatic patterns and seasonality (e.g., El Niño-Oscilación del Sur - ENSO, droughts, etc.), such as those projected for 2050, may influence the IAS Cichlids’ establishment and spread in the study area73,74,75,76. Several authors also demonstrated that in a large regional-scale study, simple influences, such as altitude and basin area, seem enough to identify the distribution and suitable habitats for freshwater fish species77,78. In our case, precipitation and temperature seasonality were the main drivers of habitat suitability changes.
The contrasting habitat suitability patterns observed among introduced IAS Cichlids reflect their species-specific evolutionary histories, ecological requirements, physiological tolerances, and the establishment time in the basin72. Cichla spp. native to the adjacent Araguaia-Tocantins basin (contiguous to Br-UPRB) showed greater suitability in the Pantanal wetland, exhibiting preference for warm, shallow lentic or slow-flowing environments with complex structural habitats characteristic of floodplain ecosystems in both basins (native and introduced)79. Despite flowing into opposite mouths, the Amazon and Paraná rivers not only share similar environmental conditions but also share related freshwater fauna80. These antecedents could be related to the successful establishment of Cichla spp. in the Br-UPRB and other tropical ecoregions of the Americas, Africa, Southeast Asia, and the Pacific islands81,82,83. On the other hand, O. niloticus shows greater suitability in the southernmost tributary sub-basins of the Brazilian plateau, which provide more stable thermal conditions and diverse habitat heterogeneity that align with this species’ broader environmental tolerance, opportunistic feeding behavior, and ability to thrive across a wide range of pH, conductivity, and turbidity conditions84,85. This invasion pattern has been documented in tropical and subtropical basins worldwide, including establishing the Nile Tilapia in the Pearl River system (China)86, Lake Victoria (Central Africa)87, various basins in Mozambique and Southern Africa88,89, India90,91, and several water bodies across the Americas35, where these introductions have consistently resulted in substantial ecological and economic impacts37,92,93.
Our results also support the growing evidence that climate change impacts on habitat suitability are species-specific, driven by habitat preferences, physiological variability, and thermal plasticity45,51,94. For instance, although Cichla spp. use habitats similarly; our ENMs show that the most suitable areas in the basin are exclusive to each species81,83. Specifically, 80% of the suitable habitats are occupied by a single species. This suggests that under climate change, IAS Cichlids in the Br-UPRB may not compete with each other for suitable habitats, but they could instead compete with native species23. In the case of Cichla spp., our projections of suitable areas for both species are consistent with their reported areas of introduction. While C. piquiti was introduced in the early 1980s to the Piquiri River and is well-established in the floodplain of the Paraguay River28,29, C. kelberi was introduced most recently in the Padre Inácio and Caramujo rivers (northern of Br-UPRB) and has been expanding its invasion range toward the southern basin30. Climate change projections indicate that C. piquiti may experience the most considerable expansion with an IAS Range Change of 248.1%, while C. kelberi shows a more moderate but still substantial rate of change at 56.7%. Both Cichlid species are very aggressive piscivores, and if these IAS establish viable populations in all suitable habitats along the basin where our models predicted, these species could drastically impact the native fish species populations81,83.
The previous hypothesis that the invasion range of C. piquiti was restricted by the turbid waters, operating as an ecological barrier to its spread, has recently been refuted29,95,96. Our models’ predictions include sub-basins and main rivers with turbid waters as suitable habitats for this species in the Br-UPRB, where the IAS has been fished regularly (12,961 kg in 2009, 994 kg in 2010, and 16,591 kg in 2018)31,97. The 2010 decline in catches could be related to cold fronts in July and August (4.6–9.2 °C) since this species prefers warmer waters (physiological optimum above 15 °C)98,99. Future climate scenarios with a temperature increase of at least 1.5 °C would likely favor the proliferation of this IAS. This is particularly concerning given that these IAS have already caused severe ecological damage in other ecosystems where they have been introduced, including the collapse of native fish assemblages in Brazilian reservoirs such as Três Marias, Sobradinho, and Itumbiara, the displacement of native piscivores in Colombian rivers and reservoirs, and the disruption of food webs in Venezuelan and Panamanian freshwater systems81,99,100.
Despite O. niloticus’s well-documented history as an invasive competitor, it may show stability in its future habitat-suitability with an expected IAS Range Change of 4% 35,101. However, considering the wide range and global occurrence availability, our results are limited by the number of records and the considerable geographical distance between the northernmost and southernmost occurrences of O. niloticus in the study basin. This limitation reflects the recent establishment of O. niloticus populations in the Br-UPRB and the concentrated sampling efforts in specific introduction areas rather than basin-wide and native surveys. This disjunct invasion pattern suggests the possibility of multiple events of introduction and diverse routes of spread within the study area. Given their recent economic importance in the basin, this potential IAS necessitates further research and urgent management actions35.
Contrary to our expectations, reservoirs in the basin appear to offer limited habitats for introducing IAS Cichlids, with only 15.52% of reservoir habitats currently exhibiting suitability for these species, which is substantially below the 58.10% suitability observed in the Pantanal wetland (Tables 3 and 4). Nevertheless, our projections for 2050 show that the extent of suitable habitats for these species is expected to more than double within all reservoirs, with large reservoirs projected to reach 57.14% suitability and overall reservoir suitability increasing to 34.48%. This expected habitat suitability in large reservoirs, where there are usually more existing alien fish farms, and more in the planning stages102,103. The proliferation of IAS Cichlids in Br-UPRB reservoirs may also be determined by native species interaction, including predation, competition, and upstream ecosystems, as has been happening in other tropical reservoirs26,99,104,105.
We compiled information on public datasets built from traditional scientific activity (field and laboratory occurrence records, online and museum databases), supplemented by data from Citizen Science (Structured interviews and Instagram posts). We highlight the recent use of fishers’ local ecological knowledge (LEK) as a valuable alternative for gathering information on Cichla spp.106. Of all the occurrences documented in the study area through this research, 74.4% were contributed by Citizen Science Data. This approach is not always easy to implement, some fishers were reluctant to participate in interviews and share information on fishing locations, primarily due to their unfamiliarity with our research or apprehension regarding fishing regulations, including conservation areas and spawning season, regardless of the IAS status37,107. Despite the drawbacks, the simplicity of visually identifying the IAS Cichlids through the fishers and photographs has been another advantage of the methodology used108. However, there are challenges in visually identifying several unknown species with confidence (i.e., cryptic species). Using Citizen Data with the ENMs’ capacity to predict the regions at risk of IAS invasion is a promising and cost-effective method to develop practical preventative actions and planning management in the basin109,110.
Conclusion and recommendations
Previous research on IAS Cichlids introduced in the basin was based primarily on the fishing catch reports, focusing on a single species within a specific area or sub-basin29,30,31. Our study provides a comprehensive Spatio-temporal perspective on the IAS Cichlids’ potential invasive areas, covering inter-regions and intra-basin levels. The model outputs shown here present a state-of-the-art estimation of suitable habitats for the three main fish species introduced in the Br-UPRB. It represents a second step in understanding the potential IAS impacts on Pantanal’s native freshwater fish community, one of the most diverse in the Neotropical region (~ 300 species).
Mato Grosso state (northern half of Br-UPRB) restricts fishing, transport, and storage of introduced Cichla spp. through 2028 under State law 12.434111. This law, in addition to 16 other regulations protecting invasive Peacock bass (Cichla spp.) in Brazil, contributes to biodiversity loss, disrupts freshwater ecosystems, and generates social conflicts37,112. Considering the economic costs of biological invasions in the Neotropical basins113 and the unintended consequences of valuing non-native freshwater species37, we recommend the implementation of better control and management protocols in fish farm facilities and reservoirs to avoid the escapes of additional IAS fishes in suitable areas114.
As suggested by other researchers, we recommend conducting field validation of habitat-suitability models for IAS115. This involves testing whether these models adequately predict observed data and, if accurate, using them to estimate the potential range of the invasive species116. Field reports conducted in 202131, after our field data collection (2018–2019), confirm the presence of C. piquiti in several habitats that were predicted as suitable in our models. We also believe that the constant update of the models with new data sources, through some automation processes by machine learning algorithms and artificial intelligence (AI), could produce dynamic and robust models that allow a better understanding of the invasion patterns and IAS ecology117,118.
Among other actions, we also advocate for expanding the IAS impact studies to the Bolivian and Paraguayan share of the UPRB and increasing the sampling of O. niloticus, given its high-risk invasive potential. This would lead to developing joint monitoring and control plans with local fisher communities, government agencies, universities, research institutes, and recreational fishing tourism companies, e.g., by encouraging massive recreational and artisanal fishing and consumption, intending to reduce populations as a strategy to control the spread of IAS fish species119. We also promote a gradual change to the commercial production (fish farms) of native species and their repopulation into areas where they have been affected by the IAS Cichlids, safeguarding ecosystem services generated by native freshwater fishes37. The early detection of a potential IAS and the ability to map its invasion range are fundamental to effective management decisions60.
Our study also demonstrates the advantage of integrating LEK with ENMs to enhance IAS detection106,120. This combined approach improves the identification of suitable habitats, enabling precise predictions of potential invadible areas46. Integrating Citizen Science Data with GIS methodologies is critical for early detection and control, as many IAS may not yet occupy all predicted suitable habitats108,121,122. The implementation of the methods used in this study allows the development of targeted management strategies that can effectively mitigate ecological costs during the crucial initial invasion stages123. Furthermore, the integration of real-time monitoring systems with predictive modeling provides decision-makers with the tools necessary to implement rapid response protocols, prioritize conservation efforts in high-risk areas, and allocate resources efficiently before IAS populations become established and irreversibly alter native freshwater ecosystems. This approach is particularly critical in biodiversity hotspots, where prevention costs are significantly lower than post-establishment control measures.
Materials and methods
Study area
The study is focused on the Br-UPRB, or Paraguay River, and its main upstream tributaries of the Pantanal wetland in the Brazilian part101 (Fig. 2). This Ramsar site begins in the Brazilian plateau, covering an area of 361,666 km276. The basin experiences a tropical humid climate, averaging an annual temperature of 22.5 to 26.5 °C. Seasonal precipitation, mainly during the rainy period (October-March), contributes 70% of the total annual precipitation (800–1600 mm). These climatic dynamics impact substantial water level fluctuations of approximately 3.1 ± 0.9 m in the floodplain76. Over decades, the Br-UPRB has faced deforestation, ecosystem homogenization, habitat loss, fragmentation, biodiversity decline, and the introduction of IAS, all impacting regional wildlife persistence24,101.
Study area - Brazilian part of the Upper Paraguay River Basin (Br-UPRB). (1) Pantanal wetland and the main tributary sub-basins in the Brazilian plateau (2–8). (2) Jaurú-Sepotuba, (3) Cuiabá, (4) São Lourenço-Vermelho, (5) Taquari-Coxim, (6) Aquidauana-Negro, (7) Miranda, and (8) Apa-Perdido. The location of reservoirs and Introduced Alien Species (IAS) occurrences are included. The map was created using ArcGIS Pro 2.8: https://www.esri.com/en-us/arcgis/products/arcgis-pro/overview.
Mapping the hydroelectric reservoirs in the basin
To assess if the hydroelectric reservoir conditions are suitable for the IAS fishes introduced to the basin, we used the Brazilian Electric Energy Agency’s (ANEEL) georeferenced database and Landsat 8 OLI- Operational Land Imager Collection-2 Level-2 (01/01/2018-06/01/2019) satellite imagery to confirm the presence of reservoirs124,125,126. We selected reservoirs with an area over 1 km2 to match the environmental variables’ spatial resolution, and the ANEEL classification of reservoir types was followed (Table 5; Fig. 2).
Species occurrence records
We collected data on IAS Cichlid occurrences (1989–2019) from museum specimens, field catches, and databases in native and study areas (Table S8). We acquired reliable occurrences of Blue Peacock Bass: C. piquiti, Yellow Peacock Bass: C. kelberi, and Tilapia: O. niloticus in the Br-UPRB. Record gaps were filled through Citizen Science data, obtaining information from interviews with local fishers and georeferenced photos of sport anglers obtained on the social network Instagram (Meta Platforms, Inc.)111.
The semi-structured interviews with local fishers were done following the Rapid Assessment Protocol (RAP) developed for detecting IAS fish species in Brazilian lakes108. We conducted 257 interviews with experienced local fishers between 2018 and 2019, covering most of the study area (Main rivers and sub-basins). To ensure accuracy, cross-validation of interviews involved presenting images of common fish species absent in the basin and avoiding false-positive occurrence records (model of the interview in Table S9 and Figure S10). Each validated interview was transformed into an occurrence of IAS only if the fisher could specify the nearest water body section where the species was found.
Beyond interviews, an Instagram IAS georeferenced photo analysis was conducted in the study area. Original posts, official accounts, and fishing-related hashtags were cataloged. We communicated with photographers and fishers to confirm the location details. Out of 326 potential IAS records, only 80 were validated through direct messages, ensuring precise georeferencing. IAS occurrences within 1 km of the next water body were relocated to align with gridded environmental variables127. Records that did not meet this criterion or duplicated within the same 1 km² cell were eliminated, retaining only the most recent occurrence.
Environmental variables and climate scenarios
To model the current and future suitable habitats for IAS Cichlid species, we selected topographic and bioclimatic variables relevant to climate change and IAS modeling in 30 arc-seconds of resolution (≈ 1km2. For current conditions, we used the database of freshwater-specific environmental variables for biodiversity analysis127. For the future climatic scenario, we employed WorldClim bioclimatic variables from CMIP6 (Coupled Model Intercomparison Project) simulations representing 2050 conditions (average for 2041–2060) in the Shared Socio-economic Pathways SSP2 4.5, a mid-term scenario58. The SSP2 4.5 scenario foresees a moderate path for climate action, aligning with countries’ emissions commitments67. This outlook estimates a global temperature rise of 2.1–3.5 °C from 1950, with a population of 9.6 billion. The radiative forcing in the SSP2 4.5 scenario peaks at 4.5 Wm− 2 by 2050, then decreases if priority is given to international cooperation for environmental aims; nevertheless, inequalities persist58.
To reduce the uncertainty associated with a single global climate model, we used an ensemble (unweighted mean) of CMCC-ESM2 (Centro Euro-Mediterraneo sui Cambiamenti Climatici Earth System Model version 2) and HadGEM3-GC31-LL (Hadley Centre Global Environment Model version 3, Global Coupled configuration 3.1, Low-Resolution) projections128,129. This approach provides more robust predictions and better represents the potential future scenarios130,131. Additionally, it improves accuracy by capturing different climate sensitivities and regional patterns132.
We divided the climatic and topographic predictor variables into two sets: (1) for Br-UPRB + Araguaia - Tocantins basins (for modeling the two Cichla spp. habitat suitability) and (2) for Br-UPRB + Nile basins (for modeling O. niloticus habitat suitability). We utilized the variance inflation factor (VIF) to address multicollinearity, selecting the bioclimatic variables with VIF values < 5 (Tables S1-S4)133. Additionally, we included elevation, slope, and flow accumulation in models, considering their impact on IAS fish habitat selection and dispersal134. These climatic data, used in various ENMs, are considered standard for modeling freshwater species distributions77,135.
Modeling assessment and ensemble forecasting approach
Using Ecological Niche Models (ENMs), we identifed the suitable habitats for the selected IAS Cichlid species in their native range (calibration area) and introduced region (Br-UPRB) under present and future climatic conditions (Figures S11-S13)120,136. We modeled the geographic range (native + UPRB areas) to capture each IAS’s filled environmental niche breadth, avoiding underestimating potential suitable habitats when modeling only invaded ranges46,136. IAS Cichlid occurrences and environmental variables were linked using three ENM algorithms: Generalized Linear Models (GLM), Random Forest (RF), and Maximum Entropy (MaxEnt)137. These three algorithms’ ensemble forecast (EF) enhances individual projections for IAS with limited occurrence points in undersampled regions46,138. Unlike maximizing the projected range, this EF approach relies on overlapping regions of employed ENMs, which is favorable for dealing with uncertainty and species with changing niche tolerances137,139. Models were fitted using the ensemble platform for species distribution modeling “Biomod2” in the R environment140.
The three selected algorithms use different input data: GLM uses presence-absence data, while RF and MaxEnt use presence-background data141. Pseudo-absence data were generated randomly from the background outside the occurrence points, with 1,000 pseudo-absence points generated per run. To prevent model bias during training, occurrence and pseudo-absence points were weighted equally (50% total weight for presences, 50% for pseudo-absences) using the bootstrap technique142. We used stepwise feature selection with quadratic terms based on the Akaike Information Criterion (AIC) for GLM. RF was applied with 1,000 trees and the default node size. MaxEnt was conducted with 1,000 maximum iterations and a 10− 5 convergence threshold143. Species occurrences were randomly divided into 80% for model calibration and 20% for testing the models140. We assessed model performance through random cross-validation, measuring sensitivity (accuracy in identifying occurrences), specificity (accuracy in identifying background points), AUC (area-under-the-curve), and TSS (True Skill Statistic). Only individual algorithms with AUC > 0.7 and TSS > 0.5 were selected for each ensemble model.
A minimum of 100 ensemble models were constructed per species per climatic scenario by merging algorithms (“mergedAlgo” parameter), combining in each EF predictions from three algorithms selected: RF, GLM, and MaxEnt, using “EMmean” (probability mean averaging) and “EMca” (committee averaging) parameters. Only EFs with TSS > 0.7 were selected for final modeling, which may reduce the number of models contributing to the final predictions depending on individual ensemble performance.
IAS binary presence maps
We converted ENMs’ continuous frequencies into binary projections, transforming probability values (0-100) into binary (1: Predicted suitable habitat / 0: Predicted unsuitable habitat). Applying the TSS > 0.7 threshold, we selected values (70–100) as suitable habitats (conservative predictions)139. This scenario reasonably forecasts the IAS potential range with few occurrences, as in our study, which is widely used in invasion ecology 46. Suitable / Unsuitable values were extracted for each sub-basin and reservoir, analyzing IAS habitat-suitability patterns in the study area.
Analysis of climate change scenarios
We utilized the R package " BIOMOD_RangeSize” for a climate scenario analysis of our IAS overlap140. Comparing binary projections (current vs. 2050 SSP2 4.5), we assessed changes in four habitat types:
-
(a)
Areas that may no longer be suitable: predicted to be lost.
-
(b)
Areas that may remain unsuitable: predicted to remain stable.
-
(c)
Areas that may gain suitability: predicted to be gained.
-
(d)
Areas that may remain suitable: predicted to remain stable.
Model outputs have a 0.926 km2 resolution (pixel size at equator latitude). Nevertheless, the study area’s size and distance from the equator alter potential suitable habitat values, presenting results in (≈ km²) rather than exact values (Supplementary Material 1, Tables S10-S11).
Data availability
The occurrence data of the introduced Cichlid species studied in its native distribution area were obtained from available biodiversity databases and georeferenced research papers (Supplementary Material), and the occurrence data in its introduced range are available in the Figshare repository: https://doi.org/10.6084/m9.figshare.26961754. The climatic and topographic data used are available at https://www.earthenv.org/streams and https://www.worldclim.org/data/bioclim.html. Additionally, upon request, the corresponding author could provide the high-resolution rasters generated in this study.
Code availability
The script codes generated in this study have been deposited in the following Figshare repository: https://doi.org/10.6084/m9.figshare.26961754.
References
Mainka, S. A. & Howard, G. W. Climate change and invasive species: double jeopardy. Integr. Zool. 5, 102–111 (2010).
Pyšek, P. et al. Scientists’ warning on invasive alien species. Biol. Rev. 95, 1511–1534 (2020).
Weiskopf, S. R. et al. Climate change effects on biodiversity, ecosystems, ecosystem services, and natural resource management in the united States. Sci. Total Environ. 733, 137782 (2020).
Bellard, C., Cassey, P. & Blackburn, T. M. Alien species as a driver of recent extinctions. Biol. Lett. 12, 20150623 (2016).
Roy, H. E., Pauchard, A., Stoett, P. & Renard Truong, T. IPBES Invasive Alien Species Assessment: Full Report. https://doi.org/10.5281/zenodo.7430682 (2024).
IBPES. Summary for Policymakers of the Thematic Assessment Report on Invasive Alien Species and Their Control of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. https://doi.org/10.5281/ZENODO.8314303 (2023).
Pecl, G. T. et al. Biodiversity redistribution under climate change: impacts on ecosystems and human well-being. Science 355, eaai9214 (2017).
Bellard, C., Jeschke, J. M., Leroy, B. & Mace, G. M. Insights from modeling studies on how climate change affects invasive alien species geography. Ecol. Evol. 8, 5688–5700 (2018).
Essl, F. et al. Drivers of future alien species impacts: an expert-based assessment. Glob. Change Biol. 26, 4880–4893 (2020).
Roura-Pascual, N. et al. Alternative futures for global biological invasions. Sustain. Sci. 16, 1637–1650 (2021).
Seebens, H. et al. Projecting the continental accumulation of alien species through to 2050. Glob. Change Biol. 27, 970–982 (2021).
Burgiel, S. W., Muir, A. A., Invasive & species climate change and ecosystem-based adaptation: Addressing multiple drivers of global change global invasive species programme. https://doi.org/10.13140/2.1.1460.8161 (2010).
Kernan, M. Climate change and the impact of invasive species on aquatic ecosystems. Aquat. Ecosyst. Health Manag. 18, 321–333 (2015).
Harrison, I. et al. The freshwater biodiversity crisis. Science 362, 1369–1369 (2018).
Sayer, C. A. et al. One-quarter of freshwater fauna threatened with extinction. Nature 638, 138–145 (2025).
Castellanos, E. et al. Climate Change 2022-Impacts, Adaptation and Vulnerability Regional Factsheet Central and South America. http://hdl.handle.net/20.500.12123/12589 (2022).
Torremorell, A. et al. Current and future threats for ecological quality management of South American freshwater ecosystems. Inland. Waters. 11, 125–140 (2021).
Bernery, C. et al. Freshwater fish invasions: A comprehensive review. Annu. Rev. Ecol. Evol. Syst. 53, 427–456 (2022).
Muñoz-Mas, R. et al. Two centuries of Spatial and Temporal dynamics of freshwater fish introductions. Global Ecol. Biogeogr. 32, 1632–1644 (2023).
Vitule, J. R. S. et al. Comment on ‘Fish biodiversity and conservation in South America by Reis et al. (2016)’. J. Fish Biol. 90, 1182–1190 (2017).
Fragoso-Moura, E. N., Oporto, L. T., Maia-Barbosa, P. M. & Barbosa, F. A. R. Loss of biodiversity in a conservation unit of the Brazilian Atlantic forest: the effect of introducing non-native fish species. Braz J. Biol. 76, 18–27 (2016).
Wantzen, K. M. et al. UNESCO,. The Pantanal: How long will there be Life in the Rhythm of the Waters? In River culture: life as a dance to the rhythm of the waters (ed. Wantzen, K. M.) 497–536 https://doi.org/10.54677/DYRD7304 (2023).
Peluso, L. M., Mateus, L., Penha, J., Súarez, Y. & Lemes, P. Climate change May reduce suitable habitat for freshwater fish in a tropical watershed. Clim. Change. 176, 44 (2023).
Zanatta, S. C. & Maciel, J. Pantanal ameaçado: a construção de Represas e a insignificância Da energia hídrica produzida. Inter 333–348. https://doi.org/10.20435/inter.v22i2.2811 (2021).
Agostinho, A. A., Pelicice, F. M. & Gomes, L. C. Dams and the fish fauna of the Neotropical region: impacts and management related to diversity and fisheries. Braz J. Biol. 68, 1119–1132 (2008).
Pereira, L. S. et al. Non-native species in reservoirs: how are we doing in brazil? Hydrobiologia 817, 71–84 (2018).
Rocha, B. S., García-Berthou, E. & Cianciaruso, M. V. Non-native fishes in Brazilian freshwaters: identifying biases and gaps in ecological research. Biol. Invasions. 25, 1643–1658 (2023).
Carvalho, D. C., Oliveira, D. A. A., Sampaio, I. & Beheregaray, L. B. Analysis of propagule pressure and genetic diversity in the invasibility of a freshwater apex predator: the Peacock bass (genus Cichla). Neotrop. ichthyol. 12, 105–116 (2014).
Resende, E., Marques, D. & Ferreira Lksg. A successful case of biological invasion: the fish cichla piquiti, an Amazonian species introduced into the Pantanal, Brazil. Braz J. Biol. 68, 799–805 (2008).
Ortega, J. First record of Peacock bass cichla Kelberi Kullander & Ferreira, 2006 in the Brazilian Pantanal. BIR 4, 133–138 (2015).
Cordova Junior, R. F. & D. Introdução e distribuição do tucunaré-azul e do Tambaqui no Pantanal Corumbaense. Acta 19, 01–15 (2021).
Nobile, A. B. et al. Status and recommendations for sustainable freshwater aquaculture in Brazil. Reviews Aquaculture. 12, 1495–1517 (2020).
Moreira, E. A. & Silva, D. D. P. A dozen fish: some exotic invasive species found in Brazilian waters and their impacts. Ciênc. anim. bras. 24, e-74647E (2023).
Lima Junior, D. P. et al. Aquaculture expansion in Brazilian freshwaters against the Aichi biodiversity targets. Ambio https://doi.org/10.1007/s13280-017-1001-z (2018).
Cassemiro, F. A. S., Bailly, D., Graça, D., Agostinho, A. A. & W. J. & The invasive potential of tilapias (Osteichthyes, Cichlidae) in the Americas. Hydrobiologia 817, 133–154 (2018).
De Medinas, M. et al. Predicted impacts of proposed hydroelectric facilities on fish migration routes upstream from the Pantanal wetland (Brazil). River Res. Apps. 36, 452–464 (2020).
Pelicice, F. M. et al. Unintended consequences of valuing the contributions of non-native species: misguided conservation initiatives in a megadiverse region. Biodivers. Conserv. 32, 3915–3938 (2023).
Pelicice, F. M., Latini, J. D. & Agostinho, A. A. Fish fauna disassembly after the introduction of a voracious predator: main drivers and the role of the invader’s demography. Hydrobiologia 746, 271–283 (2015).
Elith, J. & Franklin, J. Species distribution modeling. in Encyclopedia of Biodiversity 558–572 (Elsevier, doi:https://doi.org/10.1016/B978-0-12-822562-2.00379-0. (2024).
Morin, X. & Thuiller, W. Comparing niche- and process‐based models to reduce prediction uncertainty in species range shifts under climate change. Ecology 90, 1301–1313 (2009).
Soberón, J. M. Niche and area of distribution modeling: a population ecology perspective. Ecography 33, 159–167 (2010).
Valencia-Rodríguez, D., Jiménez-Segura, L., Rogéliz, C. A. & Parra, J. L. Ecological niche modeling as an effective tool to predict the distribution of freshwater organisms: the case of the Sabaleta Brycon Henni (Eigenmann, 1913). PLoS ONE. 16, e0247876 (2021).
Soberon, J. & Peterson, A. T. Interpretation of models of fundamental ecological niches and species’ distributional areas. Biodiv. Inf. 2, (2005).
Soberón, J. M. & Nakamura, M. Niches and distributional areas: Concepts, methods, and assumptions. Proc. Natl. Acad. Sci. U S A. 106, 19644–19650 (2009).
Peterson, A. T. & Soberón, J. M. Integrating fundamental concepts of ecology, biogeography, and sampling into effective ecological niche modeling and species distribution modeling. Plant. Biosystems - Int. J. Dealing all Aspects Plant. Biology. 146, 789–796 (2012).
Elith, J. Predicting distributions of invasive species. In Invasive Species (eds Robinson, A. P., Walshe, T., Burgman, M. A. & Nunn, M.) 93–129 https://doi.org/10.1017/9781139019606.006 (Cambridge University Press, 2017).
Franklin, J. Species distribution modelling supports the study of past, present and future biogeographies. J. Biogeogr. 50, 1533–1545 (2023).
Sinclair, S. J., White, M. D. & Newell, G. R. How useful are species distribution models for managing biodiversity under future climates? E&S 15, art8 (2010).
Valencia-Rodríguez, D. et al. Comparing methods for estimating geographic ranges in freshwater fishes: several mirrors of the same reality. Freshw. Biol. 70, e70014 (2025).
Saupe, E. E. et al. Variation in niche and distribution model performance: the need for a priori assessment of key causal factors. Ecol. Model. 237–238, 11–22 (2012).
Hirzel, A. H. & Le Lay, G. Habitat suitability modelling and niche theory. J. Appl. Ecol. 45, 1372–1381 (2008).
Nori, J. & Rojas-Soto, O. On the environmental background of aquatic organisms for ecological niche modeling: a call for caution. Aquat. Ecol. 53, 595–605 (2019).
Rinaldo, A., Gatto, M. & Rodriguez-Iturbe, I. River Networks as Ecological Corridors: Species, Populations, Pathogens https://doi.org/10.1017/9781108775014 (Cambridge University Press, 2020).
Thuiller, W. Ecological niche modelling. Curr. Biol. 34, R225–R229 (2024).
Fletcher, D. H., Gillingham, P. K., Britton, J. R., Blanchet, S. & Gozlan, R. E. Predicting global invasion risks: a management tool to prevent future introductions. Sci. Rep. 6, 26316 (2016).
Srivastava, V., Lafond, V. & Griess, V. C. Species distribution models (SDM): applications, benefits and challenges in invasive species management. CABI Rev.. 1–13. https://doi.org/10.1079/PAVSNNR201914020 (2019).
IPCC. Climate Change 2021 – The Physical Science Basis: Working Group I Contribution To the Sixth Assessment Report of the Intergovernmental Panel on Climate Change https://doi.org/10.1017/9781009157896 (Cambridge University Press, 2023).
Meinshausen, M. et al. The shared socio-economic pathway (SSP) greenhouse gas concentrations and their extensions to 2500. Geosci. Model. Dev. 13, 3571–3605 (2020).
Feldman, M. J. et al. Trends and gaps in the use of citizen science derived data as input for species distribution models: A quantitative review. PLoS ONE. 16, e0234587 (2021).
Gervazoni, P. et al. Citizen science improves the known and potential distribution of a strong wetland invader: implications for niche modeling and invasion management. Environ. Manage. 71, 1176–1187 (2023).
Hallwass, G., Schiavetti, A. & Silvano, R. A. M. Fishers’ knowledge indicates Temporal changes in composition and abundance of fishing resources in Amazon protected areas. Anim. Conserv. 23, 36–47 (2020).
Howard, L., Van Rees, C. B., Dahlquist, Z., Luikart, G. & Hand, B. K. A review of invasive species reporting apps for citizen science and opportunities for innovation. NB 71, 165–188 (2022).
Phillips, T. B., Bailey, R. L., Martin, V., Faulkner-Grant, H. & Bonter, D. N. The role of citizen science in management of invasive avian species: what people think, know, and do. J. Environ. Manage. 280, 111709 (2021).
Pocock, M. J. O. et al. Citizen science is a vital partnership for invasive alien species management and research. iScience 27, 108623 (2024).
Havel, J. E., Kovalenko, K. E., Thomaz, S. M. & Amalfitano, S. Kats, L. B. Aquatic invasive species: challenges for the future. Hydrobiologia 750, 147–170 (2015).
Sultan, M. & Pei, D. S. Invasive alien species problem in the reservoir. in Reservoir Ecotoxicology 219–228 (Springer International Publishing, Cham, doi:https://doi.org/10.1007/978-3-031-26344-6_19. (2023).
Nazarenko, L. S. et al. Future Climate Change Under SSP Emission Scenarios With GISS-E2.1. J. Adv. Model Earth Syst. 14, e2021MS002871 (2022).
Peluso, L. M. et al. Climate change negative effects on the Neotropical fishery resources May be exacerbated by hydroelectric dams. Sci. Total Environ. 828, 154485 (2022).
Mofu, L. et al. Impacts of non-native fishes under a seasonal temperature gradient are forecasted using functional responses and abundances. NB 49, 57–75 (2019).
Sloan, T. J. & Turingan, R. G. The effects of temperature on the feeding performance of invasive fishes. AJLS 2, 63 (2014).
Luz, S. C. S. et al. Reduced genetic diversity and the success of the invasive Peacock bass (Cichliformes: Cichlidae). Braz J. Biol. 84, e248656 (2024).
Burress, E. D. Cichlid fishes as models of ecological diversification: patterns, mechanisms, and consequences. Hydrobiologia 748, 7–27 (2015).
Alves, J. C., Andreotti, G. F., Agostinho, A. A. & Gomes, L. C. Effects of the El Niño Southern Oscillation (ENSO) on fish assemblages in a Neotropical floodplain. Hydrobiologia 848, 1811–1823 (2021).
Andreotti, G. F., Alves, J. C., Alves, D. C., Agostinho, A. A. & Gomes, L. C. The response of fish functional diversity to the El Niño Southern Oscillation (ENSO) in a Neotropical floodplain. Hydrobiologia 848, 1207–1218 (2021).
Silva, C. B., Silva, M. E. S. & Ambrizzi, T. Climatic variability of river outflow in the Pantanal region and the influence of sea surface temperature. Theor. Appl. Climatol. 129, 97–109 (2017).
Thielen, D. et al. Quo vadis pantanal? Expected precipitation extremes and drought dynamics from changing sea surface temperature. PLoS ONE. 15, e0227437 (2020).
Barradas, J. R. S., Silva, L. G. & Harvey, B. C. Fontoura, N. F. Estimating migratory fish distribution from altitude and basin area: a case study in a large Neotropical river. Freshw. Biol. 57, 2297–2305 (2012).
Comte, L. & Grenouillet, G. Distribution shifts of freshwater fish under a variable climate: comparing climatic, bioclimatic and biotic velocities. Divers. Distrib. 21, 1014–1026 (2015).
Winemiller, K. O., Winemiller, L. C. K. & Montaña, C. G. Peacock Bass: Diversity and Natural History of Tropical Predators (Academic, 2021).
Carvalho, T. P. & Albert, J. S. The Amazon-Paraguay Divide. In Historical Biogeography of Neotropical Freshwater Fishes (ed. Albert, J.) 192–202 https://doi.org/10.1525/california/9780520268685.003.0011 (University of California Press, 2011).
Agostinho, A. A. et al. Introduced cichlids in the americas: distribution Patterns, invasion Ecology, and impacts. in The Behavior, Ecology and Evolution of Cichlid Fishes (eds (eds Abate, M. E. & Noakes, D. L. G.) 313–361 https://doi.org/10.1007/978-94-024-2080-7_10. (Springer, 2021).
Sastraprawira, S. M. et al. A review on introduced cichla spp. And emerging concerns. Heliyon 6, e05370 (2020).
Franco, A. C. S., Petry, A. C., Tavares, M. R., De Fátima Ramos Guimarães, T. & Dos Santos, L. N. Global distribution of the South American Peacock basses cichla spp. Follows human interference. Fish Fish. 23, 407–421 (2022).
Gonçalves-de-Freitas, E. et al. Social behavior and welfare in nile tilapia. Fishes 4, 23 (2019).
Abd El-Hack, M. E. et al. Effect of environmental factors on growth performance of nile tilapia (Oreochromis niloticus). Int. J. Biometeorol. 66, 2183–2194 (2022).
Shuai, F. & Li, J. Nile tilapia (Oreochromis niloticus Linnaeus, 1758) invasion caused trophic structure disruptions of fish communities in the South China river—Pearl river. Biology 11, 1665 (2022).
Omondi, A. & Merceline, A. Invasive species and their impacts on the ecology of lake Victoria: A rapid review. PASJ https://doi.org/10.47787/pasj.v3i02.22 (2024).
Mojekwu, T. O. & Hoareau, T. Mozambique tilapia oreochromis mossambicus (Peters 1852) and the threat from oreochromis niloticus (Linnaeus 1758) in South africa: A review. https://doi.org/10.5281/ZENODO.13894309 (2024).
Zengeya, T. A., Robertson, M. P., Booth, A. J. & Chimimba, C. T. A qualitative ecological risk assessment of the invasive nile tilapia, Oreochromis niloticus in a sub-tropical African river system (Limpopo River, South Africa). Aquat. Conserv. 23, 51–64 (2013).
Sandilyan, S. Alien fish species in Indian inland wetlands: current status and future challenges. Wetlands Ecol. Manage. 30, 423–437 (2022).
Mayank, P. & Dwivedi, A. Invasive potential of nile Tilapia, Oreochromis niloticus (Linnaeus, 1758) from the tributary of the Ganga River, central India. J. Ear Environ. Sci. Res. 1–5 https://doi.org/10.47363/JEESR/2021(3)152 (2021).
Leal, L. B. et al. Changes in ecosystem functions generated by fish populations after the introduction of a non-native predator (Cichla kelberi) (Perciformes: Cichlidae). Neotrop. ichthyol. 19, e210041 (2021).
Rauber, R. G. et al. Changes in fish functional diversity in a Neotropical floodplain associated with the invasive cichla spp.: a long-term perspective. Hydrobiologia https://doi.org/10.1007/s10750-024-05725-w (2024).
Comte, L. & Olden, J. D. Evolutionary and environmental determinants of freshwater fish thermal tolerance and plasticity. Glob. Change Biol. 23, 728–736 (2017).
Matias, R. et al. Estudo do tucunaré (Cichla ocellaris) em duas baías marginais do Rio Piquiri (Pantanal do Paiaguás, MS): parâmetros físico- químicos e análise histológica da pele. 5, (2001).
Marques, D. K. S. & Resende, E. K. Ocorrência do Tucunaré Cichla cf. monoculus no Pantanal. https://www.infoteca.cnptia.embrapa.br/infoteca/handle/doc/812032 (2005).
Clímaco, T. S., Soriano, B. M. A. & Catella, A. C. Mortandade do tucunaré-azul (Cichla piquiti) no Pantanal em função de frentes frias: informações preliminares. (2022).
Catella, A. C., Albuquerque, S., Campos, F. & Santos, D. Sistema de Controle da Pesca de Mato Grosso do Sul - CPESCA/MS. 54 http://www.infoteca.cnptia.embrapa.br/infoteca/handle/doc/980999 (2018).
Espínola, L. A., Minte-Vera, C. V. & Júlio, H. F. Invasibility of reservoirs in the Paraná Basin, Brazil, to cichla Kelberi Kullander and Ferreira, 2006. Biol. Invasions. 12, 1873–1888 (2010).
Carvalho, T. L., De Almeida Ferreira, E., Pelicice, F. M. & Fernandes, R. Comparative functional responses predict the predatory impact of the highly invasive fish cichla Kelberi. Hydrobiologia 848, 2203–2211 (2021).
Tomas, W. M. et al. Sustainability agenda for the Pantanal wetland: perspectives on a collaborative interface for Science, Policy, and Decision-Making. Trop. Conserv. Sci. 12, 194008291987263 (2019).
Latini, A. O. et al. Brazil’s new fish farming decree threatens freshwater conservation in South America. Biol. Conserv. 263, 109353 (2021).
Ikeda-Castrillon, S. K., Oliveira-Junior, E. S., Rossetto, O. C., Saito, C. H. & Wantzen, K. M. The pantanal: A seasonal Neotropical wetland under threat. in The Palgrave Handbook of Global Sustainability 1–27 (Springer International Publishing, Cham, doi:https://doi.org/10.1007/978-3-030-38948-2_36-1. (2022).
Agostinho, A. A., Gomes, L. C., Santos, N. C. L., Ortega, J. C. G. & Pelicice, F. M. Fish assemblages in Neotropical reservoirs: colonization patterns, impacts and management. Fish. Res. 173, 26–36 (2016).
Pelicice, F. M. & Agostinho, A. A. Fish fauna destruction after the introduction of a non-native predator (Cichla kelberi) in a Neotropical reservoir. Biol. Invasions. 11, 1789–1801 (2009).
Catelani, P. A., Petry, A. C. & Pelicice, F. M. Azevedo Matias Silvano, R. Fishers’ knowledge on the ecology, impacts and benefits of the non-native Peacock bass cichla Kelberi in a coastal river in southeastern Brazil. Ethnobio Conserv. https://doi.org/10.15451/ec2020-09-10.04-1-16 (2021).
Alves, G. H. Z., Tófoli, R. M., Message, H. J., Lima-Júnior, D. P. & Hoeinghaus, D. J. New decree promotes fish invasion in Amazon and Pantanal. Biodivers. Conserv. 27, 2449–2450 (2018).
Latini, A. O. & Petrere Júnior, M. Efficiency of rapid field methods for detecting non-native fish in Eastern Brazilian lakes. Hydrobiologia 817, 85–96 (2018).
Coxen, C. L., Frey, J. K., Carleton, S. A. & Collins, D. P. Species distribution models for a migratory bird based on citizen science and satellite tracking data. Global Ecol. Conserv. 11, 298–311 (2017).
Pocock, M. J. O., Roy, H. E., Fox, R., Ellis, W. N. & Botham, M. Citizen science and invasive alien species: predicting the detection of the oak processionary moth thaumetopoea processionea by moth recorders. Biol. Conserv. 208, 146–154 (2017).
Mendes, M. LEI No 12.434, DE 01 DE MARÇO DE 2024. 1 (2024).
Ota, R. R. et al. Brazilian fish fauna threatened by a proposal for a new law protecting invaders. Biodivers. Conserv. 28, 787–789 (2019).
Adelino, J. R. P. et al. The economic costs of biological invasions in brazil: a first assessment. NB 67, 349–374 (2021).
Shirley, E. A. & Gore, M. L. Trust in scientists and rates of noncompliance with a fisheries rule in the Brazilian Pantanal. PLoS ONE. 14, e0207973 (2019).
West, A. M., Kumar, S., Brown, C. S., Stohlgren, T. J. & Bromberg, J. Field validation of an invasive species maxent model. Ecol. Inf. 36, 126–134 (2016).
Ortega-Huerta, M. A. & Vega‐Rivera, J. H. Validating distribution models for twelve endemic bird species of tropical dry forest in Western Mexico. Ecol. Evol. 7, 7672–7686 (2017).
Hutchinson, R., He, L. & Emerson, S. Species distribution modeling of citizen science data as a classification problem with class-conditional noise. AAAI 31, (2017).
Silvestro, D., Goria, S., Sterner, T. & Antonelli, A. Improving biodiversity protection through artificial intelligence. Nat. Sustain. 5, 415–424 (2022).
Moore, A. The aquatic invaders: marine management figuring Fishermen, Fisheries, and Lionfish in the Bahamas. Cult. Anthropol. 27, 667–688 (2012).
Peterson, A. T. Predicting the geography of species’ invasions via ecological niche modeling. Q. Rev. Biol. 78, 419–433 (2003).
Mohd Dali, M. Z. et al. Detection and management of freshwater invasive alien species through environmental DNA analysis and geographic information systems: A review. Sustainability 15, 9497 (2023).
Kousteni, V. et al. Citizen scientists contributing to alien species detection: the case of fishes and mollusks in European marine waters. Ecosphere 13, e03875 (2022).
Jeschke, J. M. & Strayer, D. L. Usefulness of bioclimatic models for studying climate change and invasive species. Ann. N. Y. Acad. Sci. 1134, 1–24 (2008).
ANEEL. Database of the Brazilian Electric Energy Agency. (2023).
Goward, S. N., Masek, J. G., Williams, D. L., Irons, J. R. & Thompson, R. J. The Landsat 7 mission. Remote Sens. Environ. 78, 3–12 (2001).
Du, Z. et al. Analysis of Landsat-8 OLI imagery for land surface water mapping. Remote Sens. Lett. 5, 672–681 (2014).
Domisch, S., Amatulli, G. & Jetz, W. Near-global freshwater-specific environmental variables for biodiversity analyses in 1 Km resolution. Sci. Data. 2, 150073 (2015).
Lovato, T. et al. CMIP6 Simulations With the CMCC Earth System Model (CMCC-ESM2). J. Adv. Model Earth Syst.14, eMS002814 (2022).
Andrews, M. B. et al. Historical Simulations With HadGEM3-GC3.1 for CMIP6. J. Adv. Model Earth Syst. 12, eMS001995 (2020).
Tanimu, B. et al. Enhancing reliability in climate projections: A novel approach for selecting global climate models. Phys. Chem. Earth Parts A/B/C. 134, 103598 (2024).
Castaneda-Gonzalez, M., Poulin, A., Romero-Lopez, R. & Turcotte, R. Weighting climate models for hydrological projections: effects on contrasting hydroclimatic regions. Clim. Change. 176, 170 (2023).
Firpo, M. Â. F. et al. Assessment of CMIP6 models’ performance in simulating present-day climate in Brazil. Front. Clim. 4, 948499 (2022).
Dormann, C. F. et al. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography 36, 27–46 (2013).
Camana, M., Dala-Corte, R. B. & Becker, F. G. Relation between species richness and stream slope in riffle fish assemblages is dependent on Spatial scale. Environ. Biol. Fish. 99, 603–612 (2016).
Alvarez, F., Gerhard, P., Paiva Silva, D., Spacek Godoy, B. & Montag, L. F. D. A. Effects of different variable sets on the potential distribution of fish species in the Amazon basin. Ecol. Freshw. Fish. 29, 764–778 (2020).
Rojas-Soto, O. et al. Calibration areas in ecological niche and species distribution modelling: unravelling approaches and concepts. J. Biogeogr. 51, 1416–1428 (2024).
Diniz-Filho, J. A. F. et al. Partitioning and mapping uncertainties in ensembles of forecasts of species turnover under climate change. Ecography 32, 897–906 (2009).
Araujo, M. & New, M. Ensemble forecasting of species distributions. Trends Ecol. Evol. 22, 42–47 (2007).
Hao, T., Elith, J., Guillera-Arroita, G. & Lahoz‐Monfort, J. J. A review of evidence about use and performance of species distribution modelling ensembles like BIOMOD. Divers. Distrib. 25, 839–852 (2019).
Thuiller, W. BIOMOD – optimizing predictions of species distributions and projecting potential future shifts under global change. Glob. Change Biol. 9, 1353–1362 (2003).
Sillero, N. et al. Want to model a species niche? A step-by-step guideline on correlative ecological niche modelling. Ecol. Model. 456, 109671 (2021).
Efron, B. Bootstrap methods: another look at the jackknife. Ann Statist 7, (1979).
Elith, J. et al. A statistical explanation of maxent for ecologists: statistical explanation of maxent. Divers. Distrib. 17, 43–57 (2011).
Acknowledgements
CM-G thanks OAS, GCUB & CNPq: Organization of American States, Coimbra Group of Brazilian Universities, and Brazilian Council for Scientific Research for the master scholarship (#130674/2018-4 CNPq).TS-Sc thanks FAPEMAT (project FAPEMAT-PRO.000274/2023) and CECAV (project TCCE ICMBio-CECAV nº. 001/2023).
Funding
Brazilian National Water and Basic Sanitation Agency (ANA: Agência Nacional de Águas e Saneamento Básico) for funding the field interviews.
Author information
Authors and Affiliations
Contributions
CM-G and JMFP conceived, designed, and approved the study. LAdFM and YRS provided the data collection in the field. TS-S contributed to the development of the models and/or revisions. CM-G carried out the analysis and wrote the first draft of the manuscript. All authors contributed to subsequent versions and the interpretation of the data and results. All authors approved the final version of the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Ethics declarations
This study included the participation of local fishing communities of artisanal and recreational fishers and was approved by the Federal University of Mato Grosso’s ethical research committee (register: 3.511.327). All the interviews were performed following relevant guidelines and regulations, protecting the identity of the interviewed fishers. Additionally, we confirmed that all interviewed fishers signed informed consent.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Martínez-González, C., Mateus, L., Sobral-Souza, T. et al. Climate change may increase the suitable habitats for invasive freshwater cichlids in a Neotropical basin. Sci Rep 15, 45541 (2025). https://doi.org/10.1038/s41598-025-30425-3
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41598-025-30425-3
