Study areas and tick sampling
Tick sampling was conducted in ten distinct areas, all located in the city of Turku (Fig.
1), between late May and early July of 2017. Five of the areas (hereafter “primary study areas”) were sampled four times at weekly or biweekly intervals, whereas another five (hereafter “secondary”) areas were visited only twice. The data tabulated in Online Resource
1 describe the study area and exact study sites within the areas, as well as dates of sampling.
Two of the study areas were suburban islands (Ruissalo and Hirvensalo, which were a primary and secondary study area, respectively) that were also monitored in earlier city investigations (Mäkinen et al.
2003; Sormunen et al.
2016b,
c). Eight other study areas were located on the mainland. Urban city parks, yards and vegetation-flanked walkways in the grid-planned city centre of Turku were divided into eastern and western sides using the Aura River (Fig.
1), forming one primary study area each. One primary study area was established within the main campus area of the University of Turku, while the fifth one was close to housing estates in Koroinen/Halinen (Fig.
1). Two city parks on the eastern side of the city, Urheilupuisto and Samppalinna, were also sampled during earlier studies (Mäkinen et al.
2003; Sormunen et al.
2016b,
c).
In addition to Hirvensalo Island, the suburban areas Katariina, Kuninkoja, Littoinen and Luolavuori were visited twice during the sampling period and were thus considered as secondary areas. Although all these areas also contain walking/jogging/nature trails for recreational purposes, the primary study area Ruissalo Island is by far the most popular district of the city for many types of outdoor activities, such as walking, jogging, biking, golfing, berry-picking, bird-watching, swimming, camping, and music festivals, and it is visited by hundreds of thousands of people annually. All five primary and one secondary (Hirvensalo Island) study areas consisted of 3–5 nearby but separate study sites (Fig.
1; Online Resource
1). In the other four secondary areas, only one site was studied.
In the study sites, collection of questing ticks was conducted using a cloth-dragging method, in which a 1-m
2 white cotton cloth was dragged through ground vegetation at a slow walking pace. Ticks that attached to the cloth were counted according to their developmental stage (larva, nymph, adult male or adult female) every ten metres, and all ticks caught using this method were preserved in ethanol-filled Eppendorf tubes and stored at −20°C for further analyses. Cloth-dragging was always conducted in dry weather from 8:00–16:00. Actual 10-m dragging transects within sites were not marked and thus were not exactly the same over consecutive sampling sessions. In most instances, 200–300 m (i.e., 20–30 drags) were dragged during a sampling session (mean 270 m; range 100–600 m). According to current knowledge,
I. ricinus is the only exophilic
Ixodes species occurring in the study area. We have not previously sampled (Sormunen et al.
2016a,
b,
c,
2018)
I. persulcatus individuals or identified them from crowd-sourced material (Laaksonen et al.
2017,
2018) from the Turku region.
Laboratory work
From the 706 sampled I. ricinus, a comprehensive subsample of 449 ticks was almost immediately screened after collection (in June–July 2017) for presence of bacterial and protozoan pathogens of human or veterinary importance. This subsample included eight I. ricinus larvae, 388 nymphs, 28 adult males and 25 adult females. Most (95%) larvae from all the study areas and a random selection of nymphs from Ruissalo Island remained unscreened at that stage.
Total DNA was extracted from ticks in the first subsample using NucleoSpin® Tissue kits (Macherey-Nagel, Germany) according to the kit protocols (Rev. 13/March 2014). Extracted DNA was stored at −20°C until analysis.
Ticks in the subsample were analysed for presence of
B. burgdorferi s.l. spirochetes, including genospecies
B. afzelii,
B. burgdorferi s.s.,
B. garinii, and some unconfirmed ones. The latter group may still contain individuals from the abovementioned or other genospecies; for example,
B. valaisiana has recently been detected from tick samples collected in the Turku region (Laaksonen et al.
2018). This non-specificity results from the analysis method, which was not specifically aimed at the genospecies level (see below). In addition, the presence of
Borrelia miyamotoi (a tick-borne relapsing fever spirochete),
Anaplasma phagocytophilum (a bacterium that causes human granulocytic anaplasmosis),
Rickettsia spp. (a bacterium that causes spotted fever),
Candidatus Neoehrlichia mikurensis (a bacterium that causes neoehrlichiosis),
Bartonella spp. (a bacterium that causes many issues, for example, cat scratch disease), and
Babesia spp. (a protozoan that causes babesiosis) were analysed.
Another subsample of 157 ticks was selected in May of 2018 to analyse presence of TBE virus (TBEV). These samples included 3 larvae, 147 nymphs and 7 adults. Most (146/157) of these ticks were collected from Ruissalo Island, while the rest (11/157) were from the university’s main campus area. The presence of bacterial and protozoan pathogens was not addressed in this subsample. Ticks in the second subsample were processed using NucleoSpin 96 RNA kits and RNA/DNA buffer sets (Macherey-Nagel, Düren, Germany) according to the kit protocols (RNA Kit: Rev. 05 April 2014 and RNA/DNA buffer set: Rev. 08 May 2014). RNA was stored at −80°C and DNA was stored at −20°C until analysis.
The bacterial pathogens
B. burgdorferi s.l.,
B. miyamotoi,
A. phagocytophilum,
Rickettsia spp., and
C. N. mikurensis and the protozoan pathogens
Babesia spp. in tick samples were screened using real-time quantitative PCR (qPCR) as described before (Sormunen et al.
2016b,
c,
2018; Laaksonen et al.
2018). All DNA samples were analysed using three replicate reactions performed in 384-well plates. At least three blank water samples were used as negative controls in each assay. The samples were considered positive only when a successful amplification was detected in all three replicate reactions. Samples that were identified as positive for
Babesia and
Rickettsia were subsequently sequenced to determine the species, as described before (Laaksonen et al.
2018; Sormunen et al.
2018).
For the
B. burgdorferi s.l. genospecies identification, two genospecies-specific duplex assays were used, with one for
B. miyamotoi and
B. garinii (Bmi/Bga) and another for
B. afzelii and
B. burgdorferi s.s. (Baf/Bbss) (Online Resource
2; Tveten
2013). For Bmi/Bga, we used an 8 μL reaction volume, and it contained 4 μL of SensiFAST Probe Lo-ROX Kit (Bioline, Germany), 200 nM Bmi primer, 300 nM Bga primer, 100 nM Bmi probe, 150 nM Bga probe and 3 μL of DNA template. Assays for Baf/Bbss were likewise run in 8 μL total volumes, containing 4 μL of SensiFAST Probe Lo-ROX kit, 200 nM Baf primer, 300 nM Bbss primer, 100 nM Baf probe, 150 nM Bbss probe and 3 μL of DNA template. Samples that were not found to be positive during either duplex assay were placed in the unconfirmed category.
The thermal cycling profile used for the Bmi/Bga and Baf/Bbss assays started at 95°C for 5 min, followed by 50 cycles of 95°C for 10 s and 60°C for 30 s. Thermal cycling was performed at the Finnish Microarray and Sequencing Centre (FMSC, Turku, Finland) using a QuantStudio™ 12K Flex Real-Time PCR System (Life Technologies Inc., Carlsbad, CA, USA). All qPCR results were analysed using QuantStudio™ 12K Flex Software v.1.2.2.
Tick-borne encephalitis virus RNA was screened by real-time reverse transcription-PCR as described by Schwaiger and Cassinotti (
2003) and modified by Laaksonen et al. (
2017). All samples were run in triplicate at the FMSC using the QuantStudio™ 12K Flex Real-Time PCR System. Water samples were used as blank controls in each PCR batch. Additionally, multiple strains of TBEV RNA were used as positive controls in each batch of reactions to confirm specificity. Positive control samples emitted strong signals during each reaction.
Statistical analyses
We refrained from performing formal analyses of the differences in tick abundance among the sampling areas, because the null hypothesis regarding equal abundance is redundant to start with; no plausible biological arguments would lead us to expect the same quantity of ticks in very different types of urban and suburban biotopes.
Reasonably high numbers of samples and positive pathogen findings among the nymphs, however, enabled comparisons related to the prevalence of
B. burgdorferi s.l. (three confirmed genospecies were pooled with the positives of unconfirmed genotypes),
Rickettsia spp., and
A. phagocytophilum between island and mainland samples. We chose to pool the samples from the islands and the mainland for statistical tests because sufficiently large sample sizes were not obtained from individual areas, apart from Ruissalo Island and the university campus study areas. In addition, suburban Ruissalo and Hirvensalo Islands can be viewed as passages to the rural, more natural and endemic tick areas of the Turku Archipelago (Mäkinen et al.
2003; Sormunen et al.
2016a), enabling an interesting comparison to the highly structured areas on the mainland.
The probability that a nymph will be positive for
B. burgdorferi s.l. (or for
Rickettsia spp. or
A. phagocytophilum in separate models) was modelled using a generalized linear model (GLM) with binary error distribution and a logit link function. The island-mainland classification was the only fixed explanatory factor. No random factors were set. Model-derived, mean probability estimates with their asymmetric 95% confidence intervals are given throughout the results. The GLMs were run with the GLIMMIX procedure in SAS v. 9.4 (Stroup
2013).