Background

Factors associated with global change, including climate change, habitat alteration and the introduction of invasive species, have largely contributed to the emergence of diseases of public health relevance [1]. Some mosquito-borne pathogens causing diseases such as malaria and West Nile fever are major health concerns nowadays, causing human fatalities worldwide [2]. Different mosquito species have been introduced during the last decades in areas where they were not present before, potentially altering the transmission dynamics of vector-borne pathogens affecting humans and wildlife. In addition, the costs derived for the management of invasive mosquito populations and the damage that they may produce are relevant from an economic perspective [3].

The genus Aedes includes different mosquito species with a clear invasive character. Among them, the Asian tiger mosquito Aedes albopictus is a widespread mosquito species catalogued as one of the 100 most invasive species according to the Invasive Species Specialist Group [4]. Native from southeast Asia, this species has spread to different areas around the globe in the last decades [5]. In 1979, Albania was the first area where Ae. albopictus was detected in Europe [6] but, nowadays, this species is widely present in the continent. In Spain, Ae. albopictus was first identified in Catalonia and has progressively colonized areas to southern Spain [7]. Aedes albopictus represents a major public health concern as a competent vector of viruses, such as dengue virus, Zika virus and chikungunya virus, and parasites including Dirofilaria [8,9,10,11,12].

Avian malaria parasites are widespread haemosporidians of the genus Plasmodium naturally circulating between birds and mosquitoes. Endemic avian malaria produces deleterious effects on birds, affecting their health status, survival probability and reproductive success [13, 14]. A recent study in the UK also supported the role of avian malaria parasites as a major factor explaining the population decline of a common bird species, the house sparrow Passer domesticus [15]. More relevant is the detrimental effect of invasive avian Plasmodium parasites on native immunologically naïve host species [16]. For example, the introduction of avian Plasmodium relictum together with its vector Culex quinquefasciatus strongly contributed to the population decline of native bird species in Hawaii [16,17,18,19]. However, despite their importance for the transmission of avian Plasmodium, the study of avian malaria parasites has been mainly focused on the vertebrate hosts, while the role of mosquitoes has been comparatively less investigated.

The role of recently established invasive mosquito species as vectors of avian malaria is poorly studied. Aedes albopictus uses mammals as the main bloodmeal sources, although birds represent about 5–10% of its diet [20]. Avian species including Passer montanus, Turdus merula, and Gallus domesticus, among others, are known hosts of Ae. albopictus in the invaded areas [20,21,22], suggesting a potential role of this vector in the transmission of avian Plasmodium. This is further supported by laboratory studies where Plasmodium gallinaceum, P. fallax, and P. lophurae completed their cycle in Ae. albopictus [23,24,25], although with lower probability than in other mosquitoes species [18]. In field studies, the presence of avian Plasmodium DNA in Ae. albopictus has been reported in Japan [26] and in engorged mosquito females from Italy [21]. In a previous study in northeastern Spain, none of the Ae. albopictus pools tested positive for avian Plasmodium, while 4 out of 167 pools containing 1190 female Culex pipiens harbored parasites [21]. These studies suggest that some discrepancies exist regarding the contribution of Ae. albopictus in avian Plasmodium transmission in nature. Contrary to Ae. albopictus, the role of certain native species in Spain to transmit avian Plasmodium has been proven [23]. Culex pipiens mostly feeds on birds [21, 27], and the molecular screening of avian malaria parasites in mosquitoes revealed the key role of this species as vectors of avian Plasmodium in Europe [28, 29]. Furthermore, some studies have proven experimentally the capacity of Cx. pipiens to transmit different Plasmodium species [23, 30,31,32]. On the other hand, Culiseta longiareolata is considered a bird-bitten species [33], and its role as an avian Plasmodium vector has also been demonstrated experimentally [34].

Here, a molecular xenomonitoring approach was used to explore the potential role of the invasive Ae. albopictus compared with two common native mosquito species in the local transmission of avian malaria parasites of the genus Plasmodium in the recently invaded area of southern Spain. To do that, field-collected mosquitoes were used to molecularly identify the presence of avian Plasmodium in mosquito females grouped in pools and to estimate its prevalence in the mosquito populations, also testing the potential temporal and spatial variation on avian Plasmodium prevalence in mosquitoes in southern Spain.

Methods

Study area and mosquito sampling

Mosquitoes were trapped from May to November 2022 in five localities of the provinces of Granada (n = 4) and Malaga (n = 1), southern Spain. Sampling sites included two urban sites, two peri-urban sites and one natural site, which differ in terms of land use and human density (see Additional file 1). Specifically, the four sampling sites in Granada included: one natural area sited in a sewage station surrounded by agriculture fields (La Vega, 37°09′57.6"N; 3°37′27.6"W), one urban area in the Fuentenueva campus of the University of Granada (UGR) (37°10′50.4"N; 3°36′32.7"W) and two periurban areas in the Cartuja campus of the UGR (37°11′38.7"N; 3°35′51.2"W) and Gójar (37°06′40.2"N; 3°36′22.4"W) (Fig. 1). In addition, mosquitoes were collected in the urban site of the Bioparc zoological garden (Fuengirola, Malaga province, 36°32′16.7"N; 4°37′39.5"W) to identify the role of mosquitoes in the transmission of avian malaria parasites in an area with a high diversity of wildlife species. Habitat categories were assigned based on visual inspection of the areas and subsequently confirmed according to GIS analyses in QGIS v3.18.1 [35]. Details of this procedure are shown in Additional file 1.

Fig. 1
figure 1

Study area with the five sampled localities in the provinces of Granada (upper right points) and Malaga (lower left point) in southern Spain. Point colors correspond to the habitat category of each locality

Full size image

In each locality, trapping sessions were conducted every 2–3 weeks for 24 h approximately, except for the zoo, where traps operated from 7:00 p.m. to 10:00 a.m. (local time) to avoid the influence of visitors. Mosquitoes were collected with four traps per locality and sampling session, two Blacklight (UV)-CDC Miniature traps (Centers for Disease Control and Prevention, Atlanta, GA, USA) and two Biogents (BG)-Sentinel-2 traps (Biogents, Regensbourg, Germany), the last supplemented with dry ice as a source of CO2 and BG-Lure mosquito attractant. In each locality, two subzones were established, separated by approximately 10 to 50 m. In each subzone, one Blacklight (UV)-CDC Miniature trap and one Biogents (BG)-sentinel-2 trap were placed nearby. One additional CDC-UV trap placed in a third subzone was used in the zoological garden to maximize the number of mosquito captured in this area. Mosquitoes were frozen and transported to the laboratory on dry ice, where they were identified morphologically using available keys [35, 36] and maintained frozen (– 80ºC) until further molecular analyses. Mosquitoes without any apparent rest of blood meal in their abdomen corresponding to the same species, trapping session and locality were pooled in independent Eppendorfs. Each mosquito pool included the same number of mosquitoes (4 mosquito females) to avoid the potential effect of pool size on the probability of parasite detection.

Molecular analyses

Genomic DNA from each mosquito pool containing four mosquito females each was extracted using the DNeasy ® Blood & Tissue kit (Qiagen, Hilden, Germany). Molecular amplifications of Plasmodium DNA were conducted using the protocol developed by Hellgren et al. [37].  At least one negative control per 20 samples was included during DNA extractions and one negative control per plate for PCR reactions. Each sample was screened twice to avoid false-negative samples [38]. The presence of amplicons was verified in 2% agarose gels. Positive samples were sequenced in both directions using the facilities of STAB-VIDA (Lisbon, Portugal). Sequences were edited using Geneious v2023.2.1 and parasite lineages were identified and named using BLASTN (Basic Local Alignment Search Tool) with sequences deposited in GenBank and MalAvi [39] databases.

Statistical analyses

The dataset analyzed here includes information from 204 mosquito pools of three different species: Cx. pipiens (n = 165), Ae. albopictus (n = 23) and Cs. longiareolata (n = 16).

Two different analyses were conducted. First, a Fisher’s exact test was used to identify interspecific differences in avian Plasmodium infection frequencies among the three mosquito species. Subsequently, comparisons between pairs of mosquito species were tested adjusting P-values for type I error and applying the Bonferroni method. Second, a generalized linear model (GLM) with a binomial distributed error and a ‘logit’ link function was fitted to test whether locality and sampling session affected the parasite prevalence in Cx. pipiens pools. Analyses were restricted to Cx. pipiens pools because parasites were only detected in this species (see results). The model included the avian Plasmodium infection status (binary; 1: infected, 0: uninfected) as the dependent variable, and sampling locality (factor with 5 levels) and the sampling session (continuous) as independent variables. The sampling session was included in the model because the prevalence of avian Plasmodium is expected to increase in mosquitoes in southern Spain from spring to autumn [28]. The model also included the interaction between sampling locality and sampling session, as seasonal effects might be associated with local abiotic and biotic characteristics. A backward stepwise model selection procedure was performed, starting with the most complex model that included locality, sampling session and their interaction as explanatory variables. Next, the variables that presented a non-significant coefficient were excluded, from the highest to the lowest P-value. Statistical analyses were run in R using the packages stats [40] and rstatix [41]. Parasite prevalence for mosquito species was estimated using the perfect test and exact confidence limits according to pool size in Epitools (https://epitools.ausvet.com.au/).

Results

A total of 204 pools containing 816 mosquitoes were analyzed including the species Cx. pipiens (n = 165 mosquito pools), Ae. albopictus (n = 23 mosquito pools) and Cs. longiareolata (n = 16 mosquito pools). Thirty-one pools were positive for the presence of avian malaria parasites, all of them corresponding to Cx. pipiens. The Fisher’s exact test further supported the significant differences in the prevalence of infection among mosquito species (P = 0.007). The pairwise comparisons showed significant differences in the parasite prevalence between Cx. pipiens and Ae. albopictus (P = 0.049) but not between Cx. pipiens and Cs. longiareolata (P = 0.110) and between Ae. albopictus and Cs. longiareolata (P = 1.000). The estimated prevalence of infection in Cx. pipiens using Epitools was 0.051 (C.I.95%: 0.035–0.072). Further GLM analysis on Cx. pipiens pools revealed non-significant effects for sampling locality, sampling session and their interaction on parasite prevalence, with none of these variables being retained in the final model (all P > 0.050) (Fig. 2).

Fig. 2
figure 2

Percentage of Culex pipiens pools (y axis) that tested positive or negative for avian Plasmodium according to both sampling session (left) and locality (right). Numbers above the bars indicate the sample size for each sampling session/locality. Mosquitoes were trapped from May to November 2022, with the sampling sessions regularly distributed during this period

Full size image

Avian Plasmodium was detected in Cx. pipiens pools from all the sampling localities, although the identity of lineages and their prevalence differed among areas (Fig. 3). Overall, eight Plasmodium lineages were found in Cx. pipiens pools (Fig. 3), including Plasmodium vaughani SYAT05 (n = 13), P. relictum SGS1 (n = 7) and GRW11 (n = 2), P. matutinum LINN1 (n = 6) and Plasmodium sp. SYAT24 (n = 1), YWT4 (n = 1) and COLL1 (n = 1). Furthermore, a new lineage of Plasmodium sp. was found and named CXPIP34 (n = 1; GenBank reference: OR587915). Of them, the lineages SYAT05 and SGS1 were tentatively identified in a single Cx. pipiens pool (parasite coinfection) based on the identification of the double peaks in the chromatogram. Plasmodium lineages found in the different localities range from one in the periurban area of Cartuja to seven in the natural area of La Vega. Lineages SGS1 and SYAT05 were found in four of the five sampling localities, while the lineage LINN1 was found in three localities. Each of the lineages SYAT24, YWT4, COLL1 and CXPIP34 was only found in a single locality (Fig. 3).

Fig. 3
figure 3

Heat map of the prevalence of infection by different lineages of avian Plasmodium parasites (x axis) found in Cx. pipiens pools from five localities (y axis) in the provinces of Granada (Cartuja, Fuentenueva, La Vega and Gójar) and Málaga (Fuengirola). The text on the right represents the total number of Cx. pipiens pools tested and those positive for avian Plasmodium between brackets. *One pool was coinfected with the lineages SYAT05 and SGS1. The legend on the top indicates the match between color and prevalence, with a darker red as prevalence increases (x-axis), while the histogram represents the number of lineages (y-axis) exhibiting a given prevalence (in blue)

Full size image

Discussion

An extensive molecular screening of avian Plasmodium in wild native and invasive mosquitoes from two provinces of southern Spain was conducted. The local circulation of avian malaria parasites in the area was confirmed, with the species Cx. pipiens likely playing a central role in their transmission. Different lineages of avian Plasmodium belonging to at least three different morphospecies were recorded in this area. Contrary to the case of Cx. pipiens, further support for the apparent absence of avian Plasmodium in Ae. albopictus was provided, potentially supporting the low relevance of this invasive species in the local transmission of avian malaria parasites in the area.

Differences between mosquito species

Culex pipiens is an ornithophilic species with birds representing 69–97% of its diet [42]. This species frequently feeds on Plasmodium-infected birds, interacting with avian blood parasites under natural conditions [27]. Different laboratory studies have identified the competence of this species for the transmission of avian malaria parasites, including studies using different lineages and species of Plasmodium [23]. For example, the presence of parasite DNA in the saliva of Cx. pipiens mosquitoes exposed to P. relictum- and Plasmodium cathemerium-infected birds was previously confirmed [30, 43, 44]. Similarly, experimental studies using lineages such as P. relictum SGS1, GRW4 and GRW11 [31, 32, 45] and P. cathemerium PADOM02 [46] also confirmed the competence of Cx. pipiens mosquitoes for their transmission. In the wild, molecular screening of parasites in mosquito pools identified the presence of avian Plasmodium in Cx. pipiens from different European countries, including Spain [28, 47, 48]. Particularly in southern Spain, the prevalence of avian Plasmodium parasites has been found to range between 2–3.2% in Cx. pipiens [28, 49], a slightly lower value than that found in this study (5.1%). Overall, our results add support for the key role of Cx. pipiens mosquitoes in the transmission of avian Plasmodium under natural conditions.

The statistical analysis supported the difference in parasite prevalence between mosquito species, with a significantly higher prevalence found in Cx. pipiens than in Ae. albopictus. A similar trend was found between Cx. pipiens and Cs. longiareolata, although the lack of significance when comparing these two species is probably due to the relatively low number of Cs. longiareolata pools analyzed here, as a total absence of avian Plasmodium was found in Ae. albopictus and Cs. longiareolata pools. Although both species are able to feed on birds, potentially affecting the transmission of avian Plasmodium, mammals clearly dominate the diet of Ae. albopictus [20, 27, 49]. This mammal-bias feeding pattern may explain the low contribution of this invasive species to the transmission of avian blood parasites. In a previous study in the Minami Daito Island of Japan, authors identified the presence of avian Plasmodium DNA in a single pool of Ae. albopictus out of the 46 mosquito pools tested containing 81 mosquitoes [26]. However, in accordance with our results, additional parasite screenings in Ae. albopictus from Spain and Japan reported the absence of avian Plasmodium [50, 56]. The low susceptibility of Ae. albopictus to avian malaria could also account for our results. LaPointe et al. [18] reported that even when avian Plasmodium could complete sporogony in both Cx. quinquefasciatus and Ae. albopictus, their susceptibility differs greatly, reducing the probability of parasite detection in Ae. albopictus. Thus, both the low exposure to avian blood parasites and a reduced competence for the development of parasites in mosquitoes may explain, at least in part, the absence of Plasmodium found in Ae. albopictus. However, this may not be the case for Cs. longiareolata, which is known as a common bird-bitten species [27]. In laboratory studies, avian malaria parasites completed their development in Cs. longiareolata females [23]. Nevertheless, to our knowledge, there are no laboratory studies on the interaction between Cs. longiareolata and the avian Plasmodium lineages circulating in southern Spain, parasites which could be unable to complete their cycle in this mosquito species. Indeed, for the case of field-collected mosquitoes in other regions of Spain, previous studies revealed a total absence of avian Plasmodium in Cs. longiareolata [50, 51]. However, Mora-Rubio et al. [51] found DNA of the related Haemoproteus parasites, which may be mainly transmitted by other insect groups (i.e. Culicoides) [52]. Thus, although the small sample size for this species could partly explain the total absence of avian Plasmodium, current evidence supports the limited epidemiological relevance of Cs. longiareolata in this geographical area for the transmission of avian malaria parasites. Furthermore, the lower relative abundance of Ae. albopictus and Cs. longiareolata in relation with Cx pipiens could also help explain its lower epidemiological importance. In addition, since traps in the zoological garden operated during a shorter period than in the rest of the localities, our sampling size in this area could be affected. However, this fact is not expected to affect the conclusions obtained based on the extremely low prevalence of avian Plasmodium found in these species here and in other previous studies [50, 51].

Lack of effect of locality and seasonality

No significant differences were found in the prevalence of infection in Cx. pipiens according to the sampling locality and sampling session. This supports a similar prevalence of infections between areas despite their diverse landscape characteristics (see Additional file 1) and likely host community composition, which may alter the epidemiology of parasites circulating in the area [53]. Sampling mosquitoes in more localities with different characteristics could result in the identification of spatial differences in parasite prevalence in Cx. pipiens. However, in line with our results, Cx. pipiens show variable avian malaria prevalence in different geographical areas [28, 54, 55], with non-significant differences among sampling points or localities in the same region [56,57,58]. However, some differences were found in relation to the parasite lineages circulating in the different areas, with some lineages (e.g. SGS1 and SYAT05) being recorded in most of the sampling localities while four lineages were only found in a single locality. A different composition of bird and/or mosquito communities in these areas may explain these differences [53]. Interestingly, three lineages were found circulating in the zoological garden, which have been previously recorded infecting different bird species in southern Spain including common blackbirds [39], a common species in this locality (authors personal observation). The circulation of these lineages in the zoological garden should be relevant from a veterinary perspective, as they have been previously associated with mortality events of immunologically naïve species such as penguins [59, 60]. On the other hand, several studies showed a variation in the prevalence of avian Plasmodium infection in Cx. pipiens mosquitoes according to season [58, 61, 62], one of them also conducted in southern Spain [28]. In this case, the maximum peak of infection prevalence was reached in autumn [28, 51], which has been attributed, among other factors, to the increased abundance of immunological naïve bird individuals (yearling birds) [28, 63,64,65]. The lack of differences found here could be partly due to the fact that most of the mosquito samples were taken in summer, while comparatively a lower number of samples from spring and autumn were included in the analyses. In addition, one limitation of our study is that mosquitoes were sampled during a single year. Although similar procedures have been used in other previous studies [28, 51], interannual differences in factors such as temperature may largely determine the community of vectors [66] and, potentially, the local circulation of mosquito-borne parasites [67].

Conclusions

Culex pipiens mosquito females play a central role in the transmission of avian Plasmodium in southern Spain. By contrast, our results suggest that the invasive Ae. albopictus might play a minor role as vectors of these pathogens, probably due to differences in their blood feeding patterns, pathogen susceptibility and relative abundance. Further studies analyzing additional mosquitoes from other areas and years should be carried out to confirm the low relevance of this invasive species in the local circulation of avian pathogens. Due to the similarities in the epidemiology of avian Plasmodium with other pathogens which use birds as reservoirs and mosquitoes as vectors, such as the zoonotic West Nile virus [68], our results could shed light on the epidemiology of pathogens with public health relevance. This is especially relevant considering the local circulation of WNV in southern Spain [69], a zoonotic pathogen occasionally transmitted by mosquitoes from birds to humans, horses and other vertebrates [70].