Biological invasions by non-native mosquito species pose significant global economic and public health threats due to their high vector competence. Still, their ecological relationships with the native fauna are rarely considered. This field study investigates the association between Aedes albopictus and the native mosquito community in Hungary. We conducted field sampling from 2020 to 2023, identified 104,612 mosquitoes from 29 species captured in traps across 532 locations, and characterized species diversity as well as the abundance of native species at each trapping event. We found that, where Ae. albopictus is already present, its abundance was positively associated with the species richness and diversity of native mosquitoes even after controlling for spatial and seasonal effects. We identified several significant pair-wise associations between Ae. albopictus and native species, both in positive and negative directions. These correlative patterns may indicate either direct interaction between species (e.g., competition) or patterns of co-existence that are driven by a third variable in the background. We performed path analyses to uncover the causal relationship among environmental traits (temperature, precipitation, urbanization and distance from water surfaces), the abundance of Ae. albopictus and that of four native species depicting co-existence patterns with the invasive one. These analyses revealed that the co-occurrences of Ae. albopictus with Anopheles plumbeus and Coquillettidia richiardii were mediated by seasonal activity, indicating that invasive mosquitoes can thrive in ecological conditions that disfavor native species. However, path models for Ochlerotatus dorsalis and Oc. geniculatus suggested a more direct causal relationship with Ae. albopictus. Although we cannot exclude that interspecific competition may partly shape the distribution of invasive and native mosquitoes, environmental filtering via shared ecological preferences appears to play a dominant role in structuring co-occurrence patterns.

Aedes albopictus, BG-Sentinel mosquito trap, competitive exclusion, ecological niche, emerging infectious diseases

Biological invasion refers to the process by which non-native species spread into new environments and establish populations, and often has significant ecological, economic, and social impacts (Simberloff et al. 2013). Human activities, such as trade and travel, play an important role in introducing and spreading non-native species worldwide. Invasive species significantly alter ecological interactions by competing with natives for resources, disrupting food webs, modifying habitats, transmitting diseases, and causing genetic changes through hybridization (McMichael and Bouma 2000; Richardson 2011; Pyšek et al. 2012). Therefore, invasive species exert their influence on other native species and ecosystems primarily through their biotic interactions, which encompass predation, parasitism, interspecific competition, and ecosystem engineering. These impacts can lead to declines in native species, ecosystem instability, and a reduction in ecosystem services crucial for human well-being. Understanding the impact of invasive species is vital to preserve native biodiversity and ecosystem functioning.

An increasing number of invasive mosquito species have appeared in Europe, posing substantial and immediate risks to public health and ecosystems, with Aedes albopictus being the most abundant and threatening (Medlock et al. 2015). The establishment and expected drastic expansion of this invasive mosquito in Europe (Carlassara et al. 2024) raises urgent concerns due to their ability to transmit various pathogens, such as viruses causing chikungunya, Zika fever, yellow fever, West Nile fever, Eastern equine encephalitis, La Crosse encephalitis, St. Louis encephalitis, Japanese encephalitis, or Rift Valley fever (Hubálek 2008; Medlock et al. 2012). Mosquito-borne diseases can significantly impact human populations, ranging from mild symptoms to severe illness and even death. Furthermore, the presence of invasive mosquito species could also disrupt local ecosystems and native species interactions, but these ecological impacts are generally less studied than the epidemiological consequences (Juliano and Lounibos 2005).

Theories about the ecological interactions of invasive species often emphasize that competition plays a pivotal role in shaping species assemblages, particularly when invasive species exploit similar resources and habitats as native species. One classic outcome of such competition is the competitive exclusion principle, which states that two species competing for the same limiting resources cannot stably coexist in the same ecological niche (Hardin 1960). In such cases, one species typically outcompetes and displaces the other, leading to a negative association between the abundance or presence of the competing species (Juliano and Lounibos 2005). However, coexistence may still emerge when species exhibit niche partitioning, whereby differences in behavior, temporal activity, habitat use, or resource preference reduce the intensity of direct competition (Schoener 1974; Chesson 2000). Niche partitioning allows for overlapping geographic distributions without direct competitive displacement. In relation to invasive mosquitoes, most available studies suggest that these species can occasionally have neutral or even positive effects on native or co-occurring species, while a few works indicate that they play a role in competition. For example, Aedes japonicus may enhance oviposition site attractiveness for Culex pipiens, thereby indirectly facilitating co-occurrence through behavioral cues (Kaufman et al. 2012). Similarly, the presence of multiple larval species, including Ae. albopictus, may dilute predation pressure on native mosquitoes, such as Culex restuans, leading to increased survival in shared habitats (Kesavaraju et al. 2007). In other cases, invasive species may modify the microbial community of breeding sites in ways that influence the growth or habitat selection of other species (Murrell and Juliano 2008). Nevertheless, some studies have demonstrated direct larval competition between invasive and native mosquitoes under controlled conditions (Juliano 1998; Juliano and Lounibos 2005), indicating that such interactions may occur under specific ecological contexts. These examples underscore that community dynamics can also include indirect facilitative effects, and that not all outcomes of species introductions are necessarily antagonistic.

One of the most common approaches to infer ecological interactions between invasive and native species from field studies is the analysis of correlations in their prevalence or abundance across space and time. These co-occurrence patterns—whether positive or negative—are often used as proxies for making inferences about biotic relationships such as competition, facilitation, or neutral coexistence (Blanchet et al. 2020). However, observed co-occurrence patterns (either positive or negative) between invasive and native species in the field may also reflect shared habitat preferences or environmental filtering rather than biotic interactions. That is, species may co-occur simply because they are independently attracted to similar environmental conditions, such as urban habitats with artificial breeding sites or areas with suitable microclimates (Leibold et al. 2004; Götzenberger et al. 2012). In this context, habitat attraction or environmental affinity could lead to positive species associations, even in the absence of direct ecological interaction. Similarly, if species match their seasonal activity with the same environmental fluctuations, a positive association will exist between the abundance of these species across time (Blonder et al. 2014). On the other hand, a negative association between species’ abundance can emerge in a given location without direct interaction between them if the species favor different ecological conditions. Accordingly, if species are attracted to different environmental optimum, they will be unlikely to be trapped at the same time or at the same location, which will lead to negative co-occurrences.

Studies, using field surveillance data, such as trap-based sampling across broad geographic scales, have shown that invasive Aedes albopictus often exhibits negative spatial associations with native species, such as Culex pipiens, Ochlerotatus triseriatus, and Anopheles quadrimaculatus, in parts of North America, suggesting the possibility of competitive exclusion (Benedict et al. 2007; Rochlin et al. 2013). However, other studies have found positive or neutral co-occurrence patterns depending on habitat type and environmental context (Leisnham and Juliano 2010), highlighting the complexity of inferring causality from field data alone. Numerous laboratory and field studies have demonstrated that interspecific competition, particularly among larval stages, could affect or play a significant role in shaping the outcomes of mosquito invasions (Juliano 1998; Braks et al. 2004; Juliano and Lounibos 2005; Kesavaraju et al. 2007). Experimental microcosms and field enclosures have repeatedly shown that the invasive Aedes albopictus is a superior competitor to both the invasive Aedes aegypti and the native North American Ochlerotatus triseriatus, often leading to reduced survival or displacement of the latter under resource-limited conditions (Teng and Apperson 2000). This competitive superiority has been attributed to more efficient resource utilization, insensitivity to inhibitory cues, and, in some cases, asymmetries in mating interference and egg hatching behavior. However, field observations also highlight that coexistence frequently occurs, suggesting condition-specific competition where abiotic factors (e.g., desiccation, temperature) or biotic interactions (e.g., predation, parasite escape) mitigate competitive exclusion. For instance, native predatory mosquitoes, such as Wyeomyia spp. or Toxorhynchites spp., may reduce the establishment success of Ae. albopictus in certain habitats. Overall, while larval competition is a central mechanism in mosquito invasions, its ecological impact is often context-dependent and modulated by broader environmental and community factors (Juliano and Lounibos 2005), while substantial evidence suggests facilitative effects on native species.

In this study, we investigate the ecological relationships between Ae. albopictus and native mosquito species using a large-scale field dataset collected across Hungary based on samples from adult traps. Ae. albopictus has been introduced in Hungary in the last 15 years (Soltész and Zöldi 2017; Sáringer-Kenyeres et al. 2020), and since then, they depicted a rapid spread (Garamszegi et al. 2023a; www.mosquitosurveillance.hu). We aim to evaluate whether the observed co-occurrence patterns among species are more consistent with direct competitive interaction or environmentally driven aggregation. Negative associations between species may point to competitive exclusion, particularly when species rely on similar larval habitats or resources and their spatial overlap is lower than expected by chance. However, environmentally driven aggregation may emerge if shared environmental preferences or activity patterns generate positive associations, even in the absence of direct biotic interaction. Such environmental mediators can also lead to negative associations between species if they differ in their ecological requirements. To investigate these possibilities, we apply path modeling to our field data collected over four years. The analyses incorporate climatic variables (temperature, precipitation), spatial and seasonal effects, and habitat characteristics (urbanization, proximity to water), allowing us to evaluate the relative contribution of interspecific interactions versus shared environmental responses in shaping mosquito community structure and to test inferences about the causal mechanisms.

The main results of this study were that i) the abundance of Ae. albopictus was positively associated with mosquito species richness and the diversity of the native mosquito fauna; ii) based on patterns of observed presence/absence of species in the traps, we detected evidence for significant positive co-occurrences for 3 species pairs and for significant negative co-occurrences for 8 species; iii) the abundance of Ae. albopictus was significantly associated with environmental variables (urbanization and the interaction between temperature and precipitation); iv) when we controlled for the potentially confounding effect of environmental variables, the positive relationship between the abundance of Ae. albopictus and that of 2 native mosquito species (Oc. dorsalis and Oc. geniculatus) were still significant.

The positive associations between the abundance of Ae. albopictus and species richness and diversity of the native mosquito fauna likely reflect shared environmental preferences rather than indicating facilitative biotic interactions. Most hypotheses about the effect of invasive species on native species emphasize the role of competition. Accordingly, competitive interactions among species should cause a negative relationship between the abundance of the invasive species and the diversity indices of the native community (e.g., Vilà et al. 2011; Mollot et al. 2017; Süle et al. 2023). However, while the population growth of an invasive species is unlikely to promote an increase in local species richness directly, it may, in some cases, contribute to greater functional diversity, spatial heterogeneity, or the creation of novel ecological opportunities within communities via indirect effects (Pyšek et al. 2012; Schirmel et al. 2016). Therefore, it is more likely that Ae. albopictus tends to colonize sites that are already favorable for mosquitoes in general—such as warm, urbanized habitats with abundant container-type breeding sites—where overall mosquito abundance and diversity are naturally and inherently higher (Juliano and Lounibos 2005; Leisnham and Juliano 2012). Thus, the positive associations may emerge because Ae. albopictus is more likely to occur where environmental conditions support a rich native community, rather than because it enhances that diversity. This interpretation aligns with the idea that environmental filtering, rather than direct interactions, plays a key role in structuring mosquito assemblages at the landscape scale.

The co-occurrence analysis presented in this study demonstrates how diversity patterns of mosquito assemblages can be potentially influenced by interactions with particular species. By testing all possible invasive–native species pairs against null expectations, we were able to identify which particular associations may be responsible for the detected positive associations between the abundance of the invasive species and the diversity of the native mosquito community. Out of the 26 native species assessed, 11 showed statistically significant associations with Ae. albopictus, including both positive and negative relationships. This relatively high proportion suggests that co-occurrence with Ae. albopictus is not random and warrants further ecological interpretation. Patterns of co-occurrences have been reported in other regions as well. For instance, Ae. albopictus was found to co-occur with native Ochlerotatus triseriatus in the eastern United States, though laboratory studies indicated a competitive advantage of the invader (Livdahl and Willey 1991; Novak et al. 1993). In South America, Ae. albopictus and Ae. aegypti frequently co-occur, despite competitive interactions demonstrated under experimental conditions (Braks et al. 2003; Braks et al. 2004). These findings, together with our results, suggest that invasive Ae. albopictus can persist in communities with native mosquitoes. However, such co-occurrence patterns do not have implications about the potential causal effects of the invasive species on native ones: the co-occurrence patterns may be rather shaped by a combination of ecological overlap, environmental filtering, and potentially biotic interactions.

To disentangle these potential mechanisms, we applied path analyses, aiming to separate the direct associations between the abundance of Ae. albopictus and that of native mosquito species from indirect effects that act through environmental traits. First, we identified those abiotic ecological factors that shape the abundance of Ae. albopictus. Specifically, we found that the species was more abundant in more urbanized sites, while there was also a significant interactive effect of temperature and precipitation on abundance. The latter pattern suggests that the positive association with temperature is more emphasized in rainy periods than in dry periods. This can be explained by a mechanism in which, for the enhancing effect of temperature, some optimal levels of water are required to provide suitable conditions for breeding. Similar relationships have been repeatedly shown for the same species in other studies relying on different populations or methodologies (Perrin et al. 2022; Zettle et al. 2022; Garamszegi et al. 2023b). These patterns are consistent with the known ecology of Ae. albopictus, which prefers warm climates, thrives in anthropogenic environments, as it predominantly breeds in artificial containers with breeding conditions depending on precipitation levels, rather than natural water bodies with more permanent water supply (Juliano and Lounibos 2005).

Given that Ae. albopictus is apparently responsive to certain environmental conditions, it is plausible that the detected co-occurrence patterns with native mosquitoes are casually driven by the similar or dissimilar ecological requirements of the species being compared. Accordingly, we found that the initially observed negative associations between Ae. albopictus and two native species. An. plumbeus and Cq. richiardii, disappeared when environmental variables were included in the models, and the path analyses revealed no direct relationship between Ae. albopictus and these native species. An. plumbeus is a native forest-associated mosquito that traditionally breeds in tree holes, but has recently expanded into suburban environments in parts of Europe, utilizing artificial breeding sites, such as tires (Lühken et al. 2023). In Hungary, it has occasionally been recorded from lithotelms and technotelms (Tóth 2004), suggesting a similar trend of local habitat expansion. Its ecological shift toward anthropogenic environments may increase overlap with invasive container-breeders like Ae. albopictus, raising the likelihood of competitive encounters. This convergence in habitat use could drive the observed negative association, particularly if both species depend on similar larval resources. However, the results indicate that the detected negative co-occurrence patterns likely reflect contrasting habitat preferences rather than direct interactions. For example, An. plumbeus may still predominantly breed in forested habitats in Hungary, while Ae. albopictus, is primarily associated with urbanized environments, making them co-occur rarely in the same sampling sites. Furthermore, it is also possible that these species follow different temporal dynamics (as reflected by the effect of detection date), and they have different peak activity periods along the year. Hence, the spatial and temporal segregation between these species is more plausibly explained by environmental filtering than by direct biotic exclusion.

A similar argument can be made for the indirect relationship between Ae. albopictus and Cq. richiardii. Cq. richiardii is a widespread Palearctic mosquito that typically breeds in permanent aquatic habitats rich in emergent vegetation, such as reed beds, lakeshores, marshes, and old riverbeds (Becker et al. 2010). Its larvae are largely sedentary and depend on submerged macrophytes for oxygen uptake. The distinct ecological preferences of Cq. richiardii—favoring vegetated, often rural aquatic habitats—contrast with the warm, urban, and container-based environments favored by Ae. albopictus. These differences emphasize the role of spatial segregation as a main explanation for the negative co-occurrence pattern between the two species.

The path analyses implied a more direct (positive) relationship between the abundance of Ae. albopictus and that of Oc. dorsalis and Oc. geniculatus. However, it is challenging to find a biological explanation for Ae. albopictus favoring large population sizes for these two native mosquitoes. Therefore, we suggest that the casual links between species occur due to some unmeasured environmental variables that would actually mediate the observed causal relationships. It is plausible that, for example, other components of habitat characteristics, additional climatic factors (such as winter temperature, wind) or microhabitat-level interactions, and predation pressure result these species being captured together more frequently than could be expected by chance. Accordingly, Oc. geniculatus is known to develop primarily in stable breeding sites such as tree holes in forested areas (Tóth 2004), but it has also been reported from technotelms—artificial containers—similar to those exploited by Ae. albopictus. This occasional overlap in larval habitats may contribute to the observed positive association in trap captures. While our analyses considered several key abiotic and landscape-scale factors, it is likely that other relevant drivers also influence mosquito community composition and it remains plausible that some causal effects were not captured in our statistical models (e.g., note the considerable amount of unexplained variance). It might be that future studies considering other ecological traits will be better at explaining the observed overlap between native and invasive species. Therefore, further research is needed to identify additional environmental factors that may underlie the observed co-occurrence patterns. Furthermore, as traps were positioned in areas where we expected high mosquito activity, such as near water bodies and in urban areas, we can also propose that these non-random aspects of sampling may have contributed to the co-occurrence of species. Accordingly, we cannot exclude that these species were not sampled in the full range of their environmental requirements. It is important to emphasize that our findings are based on correlative analyses, and should therefore be interpreted with caution, particularly regarding potential causal relationships between species. We also emphasize the need for targeted laboratory experiments to explore how specific characteristics of breeding habitats influence competitive interactions among mosquito larvae. We suggest that our correlational field-based findings can provide a valuable foundation for the design of such controlled experiments in the future.

While our study relies on a large, standardized dataset, we acknowledge that BG-Sentinel traps baited with CO₂ and synthetic lure may exhibit species-specific attractiveness, which could affect the catching probabilities of some mosquito taxa. Several studies have reported such biases, showing that trap efficacy can vary substantially across species due to differences in physiology, behavior, or olfactory sensitivity (Roiz et al. 2016; Amos et al. 2020; Lee et al. 2024). These differences can lead to over- or underrepresentation of certain species in trap samples, potentially influencing co-occurrence patterns and inferred associations. Accordingly, our interpretations are made with caution and in consideration of these methodological limitations we acknowledge that the composition of the mosquito community observed within the traps does not necessarily reflect the composition of the mosquito community in the wild. However, the target invasive Ae. albopictus, is known to be effectively sampled by this trap type in Hungary (Garamszegi et al. 2023a) and in other countries (Li et al. 2016; Wilke et al. 2019). According to Lühken et al. (2014), the BG-Sentinel trap is probably the best solution for general monitoring or surveillance programs of adult mosquitoes in Central Europe. We recognize that certain native species—particularly those preferring vegetated, shaded, or highly specialized larval habitats—may be under-sampled, especially if their adult females are less attracted to the lure configuration used. Most importantly, we applied the same trap type with identical bait across all sites and years, thereby the selectivity of the BG-Sentinel traps should not raise sampling bias for the hypotheses under test. Therefore, although absolute abundance values may reflect species-specific detection efficiency, comparative patterns across sites and between species pairs should still be robust, provided interpretations are made with caution regarding the composition of the mosquito community. Future studies incorporating multiple trap types, or calibrating species-specific detection probabilities using independent larval surveys, could further refine our understanding of these patterns. In addition, the observed co-occurrence patterns may also be mediated by differences in host-seeking behavior and dispersal capacity across mosquito species. Because BG-Sentinel traps target host-seeking females, their effectiveness partly depends on the proximity and composition of vertebrate host communities. Species with strong anthropophilic tendencies, such as Ae. albopictus, may be more frequently captured near human habitations, while ornithophilic or mammalophilic species might be underrepresented if their preferred hosts are scarce near trap sites. Furthermore, species differ in their ability to disperse from larval habitats to locate hosts, potentially decoupling adult presence from larval coexistence. Therefore, while adult mosquito assemblages offer valuable insights, they may not fully reflect larval-level interactions such as resource competition. This underscores the importance of integrating adult and larval surveillance to better infer ecological mechanisms underlying observed spatial patterns.

Taken together, we have shown that Ae. albopictus shows remarkable co-occurrence patterns with several native mosquitoes in Hungary, and that its local abundance can display positive relationships with estimates of the diversity of the endemic mosquito fauna. However, by considering some key ecological factors and by applying a path analysis, we were able to demonstrate – at least for two species – that the detected negative associations are unlikely to reflect direct interaction effects between populations (e.g., competition), and that environmental filtering explains species co-occurrences. These insights emphasize the importance of using environmentally informed models to avoid misattributing ecological effects and to better distinguish environmental associations from potential competitive processes. Our study also highlights that the laboratory results suggesting competition between the larvae of different species do not necessarily translate into population patterns detected in the field for reproducing generations. Overall, our results imply that shared ecological preferences can facilitate long-term coexistence even in the presence of potential competitive interactions. Such dynamics may ultimately promote the establishment and persistence of invasive mosquito species in new habitats, offering important implications for invasion biology and vector surveillance strategies. To better understand the ecological mechanisms behind co-occurrence, future studies should adopt a more comprehensive approach that focuses on a broader range of environmental variables and also integrates species-specific morphological, behavioral, and life-history traits into the analytical framework. Such integrative models would help to disentangle the relative importance of environmental filtering, niche differentiation, and biotic interactions in structuring mosquito assemblages. Considering the complexity of invasion processes, incorporating these functional attributes may provide critical insights into the mechanisms that facilitate coexistence or promote displacement among native and invasive mosquito species.

We would like to thank all colleagues of the Mosquito Monitor team (www.mosquitosurveillance.hu) who participated in the field work and assisted during taxonomic identification, namely Ágnes Klein, Tamara Szentiványi, Beáta Újvárosi. Two anonymous reviewers provided valuable comments on a previous version of the manuscript. The study was supported by funds from the Hungary’s National Research, Develop­ment and Innovation Office (K135841, PD-135143, RRF-2.3.1-21-2022-00006) and the Sustainable Development and Technologies National Programme of the Hungarian Academy of Sciences (NP2022-II-5/2022).