Questing ticks were collected by flagging the vegetation at 16 study sites in Slovakia during 1999-2019 ( Figure 1 ). Feeding ticks were collected from birds captured in Drienovec in 2019 (Šujanová et al., 2022). Chosen study locations represented different types of habitats: urban, suburban, natural and agricultural ( Supplementary Table S1 ). The research was done in western Slovakia (WS): four study sites in Bratislava (the campus of the Slovak Academy of Sciences - SAS, Železná studnička, Horský park, Podunajské Biskupice), Malacky, Záhorská Ves, Jurský Šúr, Vrbovce, Fúgelka, Trenčín and in eastern Slovakia (ES): Košice, Rozhanovce, Zádiel, Drienovec, Brzotín. Study sites at Martinské hole Mountains represent a region of northern Slovakia (NS).

Urban habitat is represented by three sites: Horský park and the campus of the SAS in Bratislava, and the castle park in the centre of Malacky town. All these parks are fenced.

Suburban habitat is represented by five sites: Železná studnička and Podunajské Biskupice (both in Bratislava), Jurský Šúr nature reserve, Trenčín and Košice. These sites are not fenced and pass into open country or forest.

Natural habitat is represented by Fúgelka, Martinské hole, Drienovec, Zádiel and Brzotín.

Agricultural habitat is represented by Vrbovce, Záhorská Ves and Rozhanovce.

Questing and several bird-feeding ticks were morphologically identified to species, life stages and sex using standard keys (Nosek and Sixl, 1972; Slovák, 2010). To confirm the identification of ticks and find out whether I. inopinatus (Estrada-Peña et al., 2014) occurs in Slovakia, the mitochondrial 16S rRNA locus and nuclear gene TROSPA were amplified and sequenced. The sequences were compared via BLASTn (Altschul et al., 1990) searches to sequences available in GenBank (Mangold et al., 1998; Noureddine et al., 2011; Norte et al., 2021). To amplify each gene, published primers and protocols were used according to Mangold et al. (1998) for 16S rRNA and Noureddine et al. (2011) for TROSPA. Obtained sequences were compared with sequences for I. ricinus and I. inopinatus mentioned in the previous research (16S rRNA, GenBank PopSet: 309318023; TROSPA, GenBank PopSet; 309318631 and 309318389) (Noureddine et al., 2011; Poli et al., 2020).

DNA was isolated from individual ticks by using the alkaline hydrolysis method (Guy and Stanek, 1991) and/or by using commercial kits: DNeasy Blood & Tissue Kit (Qiagen, Germany) and/or Macherey-Nagel NucleoSpin^®Tissue kit (Düren, Germany), following the manufacturer protocols. To confirm the presence of the tick DNA a 620-bp fragment of tick mitochondrial gene cytochrome b was amplified in randomly selected samples (Black and Roehrdanz, 1998) or total DNA was measured with a Nanodrop 2000c (Thermo Scientific, Wilmington, USA). Samples were stored at -20°C. Isolated DNA was screened for the presence of B. burgdorgferi s.l. by several PCR methods as described in previous publications (Derdáková et al., 2003; Hanincová et al., 2003a, 2003b; Taragel'ová et al., 2008; Chvostáč et al., 2018; Vaculová et al., 2019; Mtierová et al., 2020; Kazimírová et al., 2023). A positive (DNA from Borrelia-positive tick that was previously sequenced) and a negative control (Nuclease-Free Water, Qiagen, Germany instead of DNA) were used in each PCR reaction. The PCR products from the conventional PCRs were electrophoresed in a 1.5% agarose gel stained with GoodView™ Nucleic Acid stain (SBS Genetech, Beijing, China) and visualised with a UV transilluminator. The list of used PCR assays is summarised in Supplementary Table S2 . Borrelia-positive samples were further genotyped by RLB (reverse line blot) (Rijpkema et al., 1995; Hanincová et al., 2003a, 2003b; Taragel'ová et al., 2008) or by RFLP (restriction fragment length polymorphism) (Derdáková et al., 2003; and/or sequencing (Chvostáč et al., 2018; Vaculová et al., 2019; Kazimírová et al., 2023). Selected positive samples were further analysed using MLST (multilocus sequence typing) according to the protocol by Margos et al. (2008) and (Mtierová et al., 2020). Samples positive for B. garinii have been sequenced for the identification of B. bavariensis prevalence only since 2017.

Temporal changes in the prevalence and genetic variability were evaluated only at study sites that had been monitored for at least five years.

Purification of PCR products for Sanger sequencing was done by NucleoSpin Gel and PCR Clean-up (Machery-Nagel, Düren, Germany). The sequencing in both the forward and reverse direction was performed in Eurofins Genomics (Eurofins Genomics Germany GmbH, Ebersberg, Germany) and by SeqMe (SEQme s.r.o., Dobříš, Czech Republic). The complementary strands of each sequenced product were manually assembled into consensus sequences. The sequences were compared with GenBank entries using the basic local alignment search tool algorithm (www.ncbi.nlm.nih.gov/blast) and aligned with representative homologous sequences using the Clustal W implemented in the MEGA software version 11 (Tamura et al., 2021).

Within the monitored years, the collections were not uniform, therefore we statistically evaluated only multi-year collections at the same sites (5 and more collection years). Statistical differences in B. burgdorferi s.l. prevalence between study years and between four different habitat types were evaluated by t-test using updated PAST 3 system package (Hammer et al., 2001). Chi-square test was used to analyse the proportions of collected nymphs and adult ticks and the prevalence levels for significant independence. The level of significance was set at p < 0.05. Ninety-five percent confidence intervals (95% CI) for each proportion were calculated using an online calculator (Sergeant, 2018) at the website https://epitools.ausvet.com.au/ciproportion. The program outputs include the estimated proportion plus upper and lower limits of the specified confidence interval, using Wilson Score interval method (Brown et al., 2001). Correlations between prevalence of B. burgdorferi s.l. species and habitat types were evaluated by Multivariate - Principal component analysis (PCA) using updated PAST 3 system package (Hammer et al., 2001).

In total, 17,249 I. ricinus ticks, including 10,302 nymphs and 6,947 adults were analysed from all study sites.

Questing adult ticks were significantly more abundant than nymphs at two sites in Bratislava: Podunajské Biskupice (χ² = 30.70, p < 0.05), Železná studnička (χ² = 31.42, p < 0.05) and in Záhorská Ves (χ² = 13.11, p < 0.05) as well as at Martinské hole (χ² = 27.82, p < 0.05) and in Košice (χ² = 17.47, p < 0.05). On the contrary, nymphs were significantly more abundant at 10 sites: Vrbovce (χ² = 30.92, p < 0.05), Malacky (χ² = 230.12, p < 0.05), Šúr (χ² = 6.45, p < 0.05), Fúgelka (χ² = 432.77, p < 0.05), Bratislava – SAS (χ² = 435.62, p < 0.05), Bratislava – Horský park (χ² = 53.82, p < 0.05), Trenčín (χ² = 32.11, p < 0.05), Rozhanovce (χ² = 464.82, p < 0.05), Brzotín (χ² = 12.46, p < 0.05), Drienovec (χ² = 190.75, p < 0.05). At site Zádiel the nymphs/adults ratio was balanced ( Table 1 ).

Randomly selected samples of tick DNA collected at Bratislava - Železná studnička, Malacky, Vrbovce, Martinské hole, Drienovec and Košice were sequenced ( Table 2 ). Only samples with high quality sequences were included in the further analysis (n = 46). Sequenced samples from Drienovec (n = 16) represented ticks fed on birds: 3 ticks from Turdus merula (n = 2), 6 ticks from Turdus philomelos (n = 4), 3 ticks from Erithacus rubecula (n = 2), 1 tick from Fringilla coelebs (n = 1), 1 tick from Coccothraustes coccothraustes (n = 1), 1 tick from Garrulus glandarius (n = 1), 1 tick from Parus major (n = 1). Sequenced samples from other study sites were represented by ticks collected from vegetation. Morphological identification matched sequencing results, I. ricinus was confirmed in all tested samples. All our sequenced samples had 100% query coverage, E-value = 0, and 100% identity with more than one sample of Ixodes ricinus (KF197124.1 - Italy, KJ414457.1 – Belgium, GU074645.1 – France, OY973598.1) from the GenBank.

A total of 17,249 I. ricinus ticks, including 10,302 nymphs and 6,947 adults were tested for the presence of B. burgdorferi s.l. The total prevalence of infected ticks was 18.8% (3248/17249) (95% CI: 18.3-19.4), with 15.1% (1557/10302) (95% CI: 14.4-15.8) infected nymphs and 24.3% (1691/6947) (95% CI: 23.4-25.4) infected adults. The differences in total prevalence among nymphs and adult ticks were not significant (χ² = 2.15, p = 0.14). The lowest infection prevalence was detected at site Bratislava – SAS (6.6%, 95% CI: 5.6-7.8), the highest at site Záhorská Ves (33.9%, 95% CI: 30.7 – 37.2) ( Table 1 ). Significantly higher infection prevalence in nymphs than adults was recorded at Vrbovce (χ² = 7.63, p < 0.05). On the contrary, significantly higher infection prevalence in adult ticks than in nymphs was found at three sites: Záhorská Ves (χ² = 11.25, p < 0.05), Martinské hole (χ² = 6.59, p < 0.05) and Zádiel (χ² = 4.72, p < 0.05). At the other sites no significant differences were recorded.

A total of 3,158 (97.2%) PCR-positive samples were successfully genotyped. Nine species of B. burgdorferi s.l. were identified ( Table 3 ; Supplementary Figure S1 ). Borrelia afzelii (37.1%), B. garinii/bavariensis (24.7%), B. valaisiana (15.4%) were the most frequent and were present at all study sites. Borrelia afzelii was the dominant species at ten sites (Bratislava: Železná studnička, Horský park, SAS and Podunajské Biskupice, Záhorská Ves, Malacky, Vrbovce, Rozhanovce, Brzotín, Drienovská mokraď). The highest proportion of ticks harbouring this species was detected in Malacky (19.2%). Borrelia garinii/bavariensis was dominant at sites Fúgelka (18.7%) and Košice (18.0%) while B. valaisiana was the most frequent species at sites Záhorská Ves (13.0%) and Jurský Šúr (12.1%). Borrelia burgdorferi s.s. (4.1%) was found at majority of sites with the exception of Vrbovce, Jurský Šúr and Drienovská mokraď. This species was the most frequent in Košice (31.3%). Borrelia spielmanii (1.6%) appeared sporadically at three sites in Bratislava, in Malacky, Jurský Šúr, Fúgelka and at three sites in Eastern Slovakia with the dominance in Bratislava SAS and Malacky (25.5% each). Borrelia lusitaniae (12.6%) was recorded at all sites with the exception of Brzotín. This species considerably dominated in the Martinské hole area (73.7%). We confirmed the presence of B. bavariensis only in four ticks: 2 nymphs and 1 male from Malacky (0.1%) and one female from Fúgelka (0.03%).

Co-infections with two, three and/or four species were detected in 4.3% of infected ticks. Among mixed infections, B. garinii + B. valaisiana was the most frequent (56.3%) followed by B. afzelii mixed with B. valaisiana (10.4%) and B. garinii (8.9%). Prevalence of other species combinations ranged from 1.5 to 6.7%. Borrelia bissettii and Borrelia kurtenbachii (in mixed infection with B. burgdorferi s.s., B. garinii and B. valaisiana) were detected in one tick from Malacky and in one tick from Rozhanovce, respectively (Hanincová et al., 2003a; Kazimírová et al., 2023). The proportions of eleven types of detected co-infections is shown in Figure 2 .

The total prevalence differed significantly among four habitat types ( Table 1 ) (t = 9.66, p < 0.05). The highest prevalence was confirmed in the natural habitat (22.0%), the lowest in the urban habitat (13.2%).

Borrelia afzelii predominated in all habitats, but had the highest affinity to urban and agricultural habitats. Borrelia valaisiana and B. garinii/B. bavariensis were the most frequent species (33.0%) at suburban sites, which was also confirmed by PCA analysis ( Table 4 ; Figure 9 ). Borrelia lusitaniae was positively associated with the natural habitat. However, it is not possible to draw definitive conclusions based on the PCA analysis in case of B. burgdorferi s.s. and B. spielmanii. To confirm the correlation, it is necessary to obtain more samples of both agents.

Some studies indicate the presence of I. inopinatus ticks not only in dry Mediterranean regions, but also at sites in Central Europe (CE) (Chitimia-Dobler et al., 2018; Toma et al., 2021). Its occurrence was recorded in several regions of Germany (Chitimia-Dobler et al., 2017; Hauck et al., 2019), sympatric occurrence with I. ricinus in Romania, in western Austria and southeastern Germany (Chitimia-Dobler et al., 2018). Natural hosts include foxes, sheep and lizards (Estrada-Peña et al., 2014; Chitimia-Dobler et al., 2018). The species was recently recorded on migratory birds (Toma et al., 2021). On the other hand, the situation with regards to the presence of this tick species in CE is complicated (Kahl and Gray, 2023) and some studies suggest that it does not occur in CE (Rollins et al., 2023). Published data about I. inopinatus described by Estrada-Peña et al. (2014) raise the question of whether this species also occurs in Slovakia. Based on published data, we sequenced randomly selected samples of ticks that were morphologically determined as I. ricinus. Our preliminary analyses have not confirmed the presence of I. inopinatus in Slovakia.

Borrelia burgdorferi s.l. complex is the subject of many epidemiological studies in Europe (Rauter and Hartung, 2005; Strnad et al., 2017; Estrada-Peña et al., 2018). Global warming and climate change in Central Europe (CE) has an impact on the geographic distribution and seasonal activity of ticks and pathogens they carry as well as on the occurrence and population density of their hosts (Lindgren et al., 2000; Nuttall, 2022).

Long-term studies on the prevalence of B. burgdorferi s.l. are rare although monitoring of changes in the prevalence and species diversity at the same sites can provide precious information on the impact of global changes on ticks and tick-borne pathogens. In the present study, we summarised data on the infection prevalence and distribution of species of the B. burgdorferi s.l. complex in questing I. ricinus ticks from 16 study sites in Slovakia and from over 20 years (1999-2019). The large package of data shows considerable spatial variability within the B. burgdorferi s.l. complex even in such a small geographic area as Slovakia. In addition, marked temporal differences in the prevalence of infected ticks were observed at the same localities. These results are consistent with other European studies on spatial and temporal changes in Borrelia prevalence (e.g. Hönig et al., 2015; Okeyo et al., 2020; Hartemink et al., 2021; Medlock et al., 2022; Glass et al., 2023).

The abundance of I. ricinus is influenced by many factors and varies depending on the month/time of collection, as well as on changes that occur at individual sites over the years (for example, clearing of dense undergrowth in city parks, mowing of grassy areas etc.). Usually, nymphs are spread more uniformly than adult ticks and larvae (Gray et al., 2021). The relative abundance of questing I. ricinus nymphs in a three-year study from five selected central European countries was found to be associated with climatic conditions, but mainly with the Normalized Difference Vegetation Index, and did not significantly depend on land use categories (Rosà et al., 2018). In Slovakia, the highest density of nymphs was found in green urban sites, whereas by summarizing data from all studied countries and habitat types (urban, suburban, natural, agricultural), density of nymphs was highest in the natural habitats (Rosà et al., 2018). In our analysed datasets, nymphs were more abundant at ten study sites whereas adult ticks prevailed at five study sites. We recorded a higher overall infection prevalence of B. burgdorferi s.l. (18.8%) compared to the meta-analyses for Central Europe: 13.7% (Rauter and Hartung, 2005) and 12.3% (Strnad et al., 2017). The prevalence of Borrelia-infected nymphs (15.1%) in Slovakia was higher than 10.1% and 11.8% in the datasets from 1984-2003 (Rauter and Hartung, 2005) and from 2010-2016 (Strnad et al., 2017), respectively. The situation was similar for adult ticks with 24.3% positive ticks in 1999-2019 (our study) in contrast to previous studies - 18.6% (Rauter and Hartung, 2005) and 14.9% (Strnad et al., 2017). In contrast, the overall Borrelia prevalence in nymphs based on data from Slovakia, Germany and Italy for 2011-2013 was 19.3% (Rosà et al., 2018).

In our study, the lowest infection prevalence was recorded in urban areas (13.2%) while in a long-term study from the city of Hanover (Germany) the prevalence of infected ticks was as high as 25.5% (Glass et al., 2023). In comparison with urban sites of Slovakia, higher Borrelia prevalence was recorded in suburban, natural and agricultural sites which have more diverse vertebrate hosts composition (Rizzoli et al., 2014; Kazimírová et al., 2023). Spatial and temporal changes in the borrelial prevalence even within small geographical areas were reported from Slovakia, e.g., by Pangrácová et al. (2013) and Kazimírová et al. (2023). Depending on the location and habitat type, prevalence of infected ticks ranged from 1.0% to 27.8% (nymphs) and from 0% to 49.4% (adults). Total prevalence was found to range between 0% and 38.3% in urban/suburban sites, between 8.0% and 29.4% in natural sites and between 32.3% and 38.1% in agricultural sites and ecotones (see Kazimírová et al., 2023).

The distribution of B. burgdorferi s.l. species in a site depends on the presence of appropriate reservoir hosts (Wolcott et al., 2021; Steinbrink et al., 2022; Estrada-Peña et al., 2024). In total, nine species were identified in our study. Borrelia afzelii, B. garinii/bavariensis and B. valaisiana were the most frequent and were confirmed at all study sites. These results are consistent with data from other European studies (Strnad et al., 2017; Mysterud et al., 2019). In the suburban habitat, bird-associated species (B. garinii and B. valaisiana) prevailed, which may be related to the high number of songbirds in urban forest parks. Borrelia lusitaniae was confirmed in 12.8% positive samples from Slovakia, which is a higher proportion than 7% in the study of Strnad et al. (2017). This was caused by a high proportion of B. lusitaniae-positive ticks (73.7%) from a natural focus in a mountain region in Slovakia, Martinské hole Mts., that is not typical of Central Europe (Rusňáková Tarageľová et al., 2016). According to our results, this species has an affinity to the natural habitat, but the results are skewed by high number of B. lusitaniae-positive ticks from one study site only. By our previous methodology, identification of B. bavariensis was not possible, but we assume that confirmed infections with B. garinii might have included this species. The presence of B. bavariensis was confirmed in Slovakia by Hamšíková et al. (2017) and in a former study as the B. garinii ospA serotype 4 (Lenčáková et al., 2006). The rare species B. bissettii and B. kurtenbachii have been detected in Slovakia only once, and in coinfections with other Borrelia species (Hanincová et al., 2003b; Kazimírová et al., 2023).

Global changes are affecting all components of the natural foci of tick-borne diseases (Gray et al., 2009). Changes in the diversity of Borrelia species at the monitored sites in Slovakia were probably due to the influence of several factors:

As B. afzelii is associated with rodents, increasing/decreasing of the prevalence of this species may be driven by natural fluctuations in rodent population densities (Glass et al., 2023) as well as by changes in the species composition of the rodent communities. For example, high population densities of the common vole (Microtus arvalis), the major reservoir host for B. afzelii (Radzijevskaja et al., 2013), were registered Europe-wide (Jacob et al., 2020; Heroldová et al., 2021), and the striped field mouse (Apodemus agrarius) is currently expanding its range (Tulis et al., 2023). In years with low population densities of rodents, song birds being reservoirs of B. garinii and B. valaisiana may become the major feeding hosts and sources of infection for I. ricinus, especially in urban and suburban areas (Rizzoli et al., 2014). Borrelia lusitaniae dominated at Martinské hole until 2013, but since then its prevalence has been decreasing. At this location, a large area of the forest was destroyed, so the composition of tick hosts has probably changed.