Migratory birds as disseminators of ticks and the tick-borne pathogens Borrelia bacteria and tick-borne encephalitis (TBE) virus: a seasonal study at Ottenby Bird Observatory in South-eastern Sweden.

BACKGROUND: Birds can act as reservoirs of tick-borne pathogens and can also disperse pathogen-containing ticks to both nearby and remote localities. The aims of this study were to estimate tick infestation patterns on migratory birds and the prevalence of different Borrelia species and tick-borne encephalitis virus (TBEV) in ticks removed from birds in south-eastern Sweden.
METHODS: Ticks were collected from resident and migratory birds captured at the Ottenby Bird Observatory, Öland, Sweden, from March to November 2009. Ticks were molecularly identified to species, and morphologically to developmental stage, and the presence of Borrelia bacteria and TBEV was determined by quantitative real-time PCR.
RESULTS: A total of 1339 ticks in the genera Haemaphysalis, Hyalomma, and Ixodes was recorded of which I. ricinus was the most abundant species. Important tick hosts were the European robin (Erithacus rubecula), Blackbird (Turdus merula), Tree pipit (Anthus trivialis), Eurasian wren (Troglodytes troglodytes), Common redstart (Phoenicurus phoenicurus), Willow warbler (Phylloscopus trochilus), and Common whitethroat (Sylvia communis). Borrelia bacteria were detected in 25% (285/1,124) of the detached ticks available for analysis. Seven Borrelia species (B. afzelii, B. burgdorferi (s.s.), B. garinii, B. lusitaniae, B. turdi, B. valaisiana, and B. miyamotoi) were identified. B. turdi was recorded for the first time in ticks in Sweden. The number of Borrelia cells per tick ranged from 2.0 × 100 to 7.0 × 105. B. miyamotoi-containing ticks contained a significantly higher median number of Borrelia cells than B. burgdorferi (s.l.)-containing ticks. B. garinii and B. miyamotoi were the most prevalent Borrelia species in tick larvae. Larvae of I. ricinus with B. garinii were removed from seven bird species, particularly S. communis and A. trivialis, which may suggest that the larvae had contracted the Borrelia bacteria from or via these birds. Also, a high percentage of tick larvae containing B. miyamotoi was removed from E. rubecula. All ticks were negative for TBEV.
CONCLUSIONS: The results corroborate the view that the contributions of birds to human disease are substantial, particularly as blood hosts for ticks and for their short-, medium-, and long-distance dispersal. Moreover, several ground-foraging bird species appear to be important for the maintenance and dispersal of Borrelia species. The absence of TBEV in the ticks conforms to other similar studies.

Background

Several species of ticks are known to infest birds, and such birds can disperse ticks over short, medium and long distances [1]. Certain potentially human-pathogenic microorganisms are vectored by ticks. Some of these microorganisms may be harboured in both ticks and birds. Obviously, microorganism-containing vector-competent ticks, while feeding on a susceptible avian host, can transmit the microorganism to its host. Certainly, transmission of a blood-borne microorganism can also take place in the opposite direction, whereby infectious, transmission-competent birds (transmission hosts or reservoir-competent hosts) can infect feeding, susceptible ticks. During their flights birds can act as “vehicles” for the geographic spread of different types of human-pathogenic microorganisms, i.e. for microorganisms which are maintained in both birds and ticks, for microorganisms mainly or only present in ticks, and for microorganisms mainly or only present in birds. Tick-associated microorganisms, with proven, suspected, or unknown human pathogenicity which have been detected in bird-infesting ticks in Europe include Borrelia afzelii [2-5], Borrelia burgdorferi (s.s) [2, 4, 6, 7], Borrelia garinii [2–5, 7, 8], Borrelia lusitaniae [9], Borrelia turdi [3, 5, 10–12], Borrelia valaisiana [2–4, 10, 11, 13], Borrelia spielmanii [4, 5], Borrelia miyamotoi [2, 4, 14], Neoehrlichia mikurensis [4, 14–16], Rickettsia aeschlimannii [16, 17], Rickettsia africae [17], Rickettsia helvetica [14, 16, 18, 19], Rickettsia japonica [16], Rickettsia monacensis [16], Anaplasma phagocytophilum [14, 19], Babesia venatorum [16], Babesia microti [16], tick-borne encephalitis virus (TBEV) [16, 20], Alkhurma haemorrhagic fever virus (AHFV) [21], and Crimean-Congo haemorrhagic fever virus (CCHFV) [22].

Seasonal migration is common among birds breeding in northern latitudes but shows large variation between and within species in timing and direction of migration as well as distance travelled. Most migratory birds divide the trip into several migration legs, interspersed with time spent on stopover sites along the route. At suitable stopover sites, birds feed and prepare for the next leg of migration, providing opportunities for new ticks to attach and already fed ticks to detach from the birds, thereby following birds along the migratory routes and potentially creating new foci of tick-borne diseases.

The overall aim of this study is to better understand the potential role of birds as disseminators of ticks and tick-borne pathogens into Sweden during their northward, spring, and early summer migration and from Sweden during their southward, late summer and autumn migration. In particular, the present investigation was designed to determine the abundance of different tick species and tick stages infesting migratory and other birds captured at one of Sweden’s most important bird observatories. Moreover, we aimed at determining the prevalence of Borrelia and TBEV contents in the ticks infesting the birds.

Methods

Description of the sampling site, Ottenby Bird Observatory

Ottenby Bird Observatory is located on the southern point of the island of Öland in south-east Sweden. At the observatory, migratory birds have been banded since 1946, and more than 1 million birds have been caught, providing a solid knowledge base on which bird populations that pass through the area [23, 24]. Apart from funnelling migratory birds during migration, the area also functions as stopover site and breeding ground for many species of birds. In spring, the vast majority of birds caught at the observatory have just arrived after a night of migration, including passage over stretches of open water (the Baltic Sea), while in autumn, birds caught will be a mixture of birds arriving from areas to the north and east, and those that have spent some time on stopover in the area and restarting migration. The reserve in which the observatory is situated encompasses an area of about 25 km2 with shore meadows, pastures, and woodland dominated by stands of old oaks (Quercus robur) and birch (Betula sp.). The area is grazed by cattle (Bos taurus), sheep (Ovis aries) and fallow deer (Dama dama).

Bird captures and tick collections

Sampling of birds was approved by the Swedish Board of Agriculture, delegated through the Animal Research Ethics Committee in Linköping (decision 43–09). Birds were captured during the general trapping activities of staff members at the Ottenby Bird Observatory (56° 12′ N, 16° 24′ E). With the approval of the Swedish Museum of Natural History in Stockholm, Sweden, mist nets and Heligoland traps were used to capture birds during the periods 15 March–15 June and 15 July–15 November 2009 as part of the observatory’s standardized trapping scheme. Trapped birds were banded, measured, and identified to species level and, when possible, sexed and aged. At the observatory, trapping starts half an hour before sunrise and ends at 11 a.m., which means that the length of the trapping day varies with season. To achieve a representative sample of birds for this study, and to comport with other duties at the observatory, we devised a schedule wherein every second day each bird captured during the first four working hours (start 30 min before sunrise) was checked for ticks around the base of the beak and around the eyes and ears, and every other day each bird captured during the last four working hours (7–11 a.m.) was checked. Ticks were removed with forceps and photographed as described below. Ticks were stored individually in empty snap-lid tubes at − 70 °C.

Classification of bird migration categories

As noted above, the majority of birds screened for ticks in this study were caught during migration. We broadly classify the investigated species into residents, short-distance migrants, partial migrants, and long-distance migrants based on band recovery data from the observatory and Sweden as a whole [25]. The resident bird species category comprises four species—Passer domesticus, Passer montanus, Corvus monedula, and Pica pica–that are sedentary in the area and show little to no migratory propensity apart from local dispersal. The division between short- and long-distance migrants is based on distance between breeding areas and non-breeding areas, where species that mainly winter in Europe are classified as short-distance migrants and species wintering predominantly in Africa or Asia as long-distance migrants. A special category is the partial migrants, a category which here is used to denote species of which some individuals are residents, some are short-distance migrants, and some alternate between these categories depending on the severity of winter or other environmental cues. Thus, the categories used here reflect the populations of the species studied at Ottenby and would in some instance be categorized differently at other study locations in the distribution range.

Determination of developmental stage and species of the ticks

Each tick was photographed dorsally and ventrally using a digital USB microscope (Dino-Lite Long AM4013TL, AnMo Electronics Corp., Taiwan), as previously described [15]. Based on the photographs, each tick was morphologically identified to developmental stage (larva, nymph, or adult) and sex of adults. To identify the genus and species of the ticks, each specimen was analysed using molecular methods. All ticks were analysed by a PCR method targeting the tick mitochondrial 16S rRNA gene, as previously described [26].

Total nucleic acid extraction and cDNA synthesis from ticks

Collected ticks were homogenized individually by bead beating in 2 ml safe-lock microcentrifuge tubes (Eppendorf AG, Hamburg, Germany) with a 5-mm stainless steel bead (Qiagen, Hilden, Germany) in 350 μl RLT buffer (Qiagen), supplemented with 1% 2-mercaptoethanol (Sigma-Aldrich, Stockholm, Sweden) using a TissueLyser II (Qiagen) for 2 min at 25 Hz. After centrifugation at 20,000×g for 3 min, 300 μl supernatant was transferred to new microcentrifuge tubes for total nucleic acid (NA) extraction, using MagAttract® Viral RNA M48 kit (Qiagen) in a BioRobot M48 workstation (Qiagen), using a 65 μl elution volume. Each batch of 24 samples consisted of 22 ticks, one positive control [5 μl of B. burgdorferi (s.s.) B31 ATCC 35210 (108 cells/ml) and 5 μl inactivated TBEV strain K23, Encepur®, Chiron Vaccines, Marburg, Germany] and one negative control (H2O) that were extracted simultaneously. The eluted NA was reverse-transcribed to cDNA using illustra™ Ready-to-Go RT-PCR Beads kit (GE Healthcare, Amersham Place, UK). Twenty microlitres NA and 10 μl pd(N)6 random hexamer primers (0.25 μg/μl) were incubated for 5 min at 97 °C and then mixed with one RT-PCR bead dissolved in 20 μl RNase-free water. The mixture was incubated for 30 min at 42 °C, followed by 5 min at 97 °C, producing 50 μl cDNA. Pipetting was performed using a CAS1200 pipetting robot (Corbett Robotics Inc., San Francisco, CA) and incubation using a PTC 100 thermal cycler (M.J. Research, Inc., Waltham, MA).

Detection and quantification of Borrelia bacteria and TBEV by real-time PCR analyses

Two microlitres of cDNA per reaction was used in a LUX™ real-time PCR assay to detect and quantify a 131-bp-long fragment of the Borrelia spp. 16S rRNA gene, using genus-specific primers and a serial dilution of plasmid standard [27]. Two microlitres of cDNA per reaction was used in a multiplex TaqMan™ real-time PCR assay to detect and quantify all three subtypes of TBEV, using primers, probes, and a serial dilution of plasmid standard [28]. PCR reagents and templates were added to a 96-well real-time PCR plate by using a CAS1200 pipetting robot (Corbett Robotics Inc.). The real-time PCR analyses were carried out using the BioRad CFX96 Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA).

Determination of Borrelia species by conventional PCR and nucleotide sequencing

To determine Borrelia burgdorferi (s.s) species of the samples positive in the LUX real-time PCR assay, a nested conventional PCR assay using primers targeting the intergenic spacer region (IGS) between 5S and 23S rRNA genes was applied, as described in [29]. Samples that failed to produce PCR products with this assay were instead analysed with primers targeting the IGS between 16S and 23S rRNA genes to determine other Borrelia spp., as described in [30, 31]. Tick samples, positive for Borrelia spp. in the LUX real-time PCR assay, which failed to produce PCR products with the 5S-23S IGS assay and 16S-23S IGS assay, were denoted ‘untypeable’.

Sequencing of all the PCR products was performed by Macrogen Inc. (Amsterdam, The Netherlands). Chromatograms were initially edited using BioEdit software v7.0 (Tom Hall, Ibis Therapeutics, Carlsbad, CA) and sequences were examined using Basic Local Alignment Tool (BLAST). The appearance of dual peaks in chromatograms was interpreted as at least two Borrelia species in the sample and thus denoted as ‘mixed’. An additional file shows the aligned sequences (see Additional file 1).

Statistical analyses

Data were presented as medians with interquartile range (IQR) for numerical variables and as percentage for categorical variables. The numerical variables (i.e. the number of Borrelia spp. cells of different Borrelia spp. detected in the ticks) were analysed using Kruskal-Wallis test, which, if significant, was followed by Dunn’s multiple comparison test. The categorical variables (e.g. developmental stages of the tick; tick species; spring vs. autumn, etc.) were analysed using chi-square test, but when the expected frequency was < 5 in at least one of the cells of the contingency table, Fisher’s exact test with a confidence interval (CI) of 95% was used instead. Statistical analyses were performed, and graphs and figures were drawn using GraphPad Prism version 8.0.0 for Windows (GraphPad Software, San Diego, CA). P-values ≤ 0.05 were considered statistically significant.

Results

Ticks collected from birds and seasonal dynamics of infestation pattern

A total of 4601 bird individuals (4788 bird captures) of 65 species was examined for ticks at least once during the study period, i.e. 15 March–15 June and 15 July–15 November 2009, at the Ottenby Bird Observatory. A total of 749 bird individuals (759 bird captures) of 35 species was infested with at least one tick specimen (Table 1). Calculations based on bird captures showed that the monthly prevalence of birds infested with one or more ticks during the study period ranged from 4.90 to 31.5%, with the lowest and highest infestation prevalence in the second half of May and second half of September, respectively (Fig. 1).

Table 1

Bird species with tick infestation patterns, collected at the Ottenby Bird Observatory, Sweden, 2009

Bird species	Migratory categorya	No. of bird captures	No. of infested birds (%)	No. of ticks	Mean no. ticks per infested bird ± SE	Median no. ticksper infested bird(IQR)	Tick species and developmental stagec
							Ixodes spp.	I. r	I. f	H. p	H. m	ND
Accipiter nisus	PM	5	1 (20)	1	–	–		1N
Acrocephalus palustris	LM	21	3 (14)	4	1.3 ± 0.33	1 (1.00–2.00)		4N
Acrocephalus scirpaceus	LM	11	1 (9.0)	1	–	–		1N
Anthus trivialis	LM	39	26 (67)	93	3.6 ± 0.73	2 (1.00–4.25)	4L	52L, 33N				4
Carduelis cannabina	SM	15	1 (6.7)	1	–	–				1L
Carduelis flammea	SM	36	2 (5.6)	7	3.5 ± 1.50	3.5		7N
Certhia familiaris	PM	26	2 (7.7)	4	2.0 ± 0.00	2		2L				2
Cyanistes caeruleus	PM	58	4 (6.9)	7	1.6 ± 0.75	1 (1.00–3.25)		1N		4L		2
Emberiza schoeniclus	SM	28	2 (7.1)	3	1.5 ± 0.50	1.5		2N				1
Erithacus rubecula	SM	1551	368 (24)	619	1.7 ± 0.09	1 (1.00–2.00)	38L, 13N	278L, 194N, 3U	5L, 11N, 1A		1L	75
Fringilla coelebs	SM	37	6 (16)	8	1.3 ± 0.21	1 (1.00–2.00)		4L, 2N			1A	1
Fringilla montifringilla	SM	8	1 (13)	1	–	–		1L
Hippolais icterina	LM	50	2 (4.0)	2	1.0 ± 0.00	2		1L, 1N
Lanius collurio	LM	43	3 (7.0)	3	1.0 ± 0.00	1 (1.00–1.00)		3N
Luscinia luscinia	LM	18	10 (56)	29	2.9 ± 0.78	2 (1.75–3.25)		20L, 9N
Luscinia svecica	LM	28	2 (7.1)	2	1.0 ± 0.00	1	1N	1N
Motacilla alba	LM	23	1 (4.3)	3	–	–		2L		1L
Oenanthe oenanthe	LM	6	1 (17)	1	–	–				1L
Parus major	PM	44	3 (6.8)	8	2.7 ± 1.67	1 (1.00–6.00)		2N				6
Passer domesticus	R	16	1 (6.3)	1	–	–		1N
Phoenicurus ochruros	LM	10	2 (20)	2	1.0 ± 0.00	1		2N
Phoenicurus phoenicurus	LM	134	40 (30)	64	1.6 ± 0.22	1 (1.00–2.00)		29L, 32N				3
Phylloscopus collybita	SM	79	3 (3.8)	3	1.0 ± 0.00	1 (1.00–1.00)		2N				1
Phylloscopus trochilus	LM	810	38 (4.7)	57	1.5 ± 0.14	1 (1.00–2.00)	1L, 1N	19L, 30N	1L		2N	3
Prunella modularis	SM	44	16 (36)	24	1.5 ± 0.14	1 (1.00–1.00)	2N	3L, 19N
Regulus regulus	PM	319	9 (2.8)	9	1.0 ± 0.00	1 (1.00–1.00)		2L, 7N
Sturnus vulgaris	SM	14	3 (21)	3	1.0 ± 0.00	1 (1.00–1.00)		2N		1L
Sylvia atricapilla	SM	58	4 (6.9)	4	1.0 ± 0.00	1 (1.00–1.00)		2L, 1N				1
Sylvia communis	LM	123	27 (22)	49	1.8 ± 0.37	1 (1.00–2.00)		23L, 17N		4L		5
Sylvia curruca	LM	270	10 (3.7)	11	1.1 ± 0.10	1 (1.00–1.00)		7L, 2N				2
Troglodytes troglodytes	PM	254	72 (28)	120	1.7 ± 0.20	1 (1.00–2.00)	13L, 2N	59L, 31N, 1U	1L			13
Turdus iliacus	SM	29	9 (31)	23	2.6 ± 0.69	2 (1.00–4.50)		1L, 18N	1N			3
Turdus merula	PM	147	66 (45)	149	2.3 ± 0.25	1 (1.00–3.00)	6N	8L, 108N, 1U	1L, 2N			23
Turdus philomelos	SM	60	9 (15)	22	2.4 ± 0.58	2 (1.00–3.50)	2N	2L, 17N	1N
Unknownb		1	1 (100)	1	–	–		1N
Acrocephalus arundinaceus	LM	1	0	0
Acrocephalus schoenobaenus	LM	23	0	0
Anthus pratensis	SM	5	0	0
Carduelis carduelis	SM	15	0	0
Chloris chloris	SM	16	0	0
Carduelis spinus	SM	9	0	0
Carpodacus erythrinus	LM	2	0	0
Coccothraustes coccothraustes	PM	1	0	0
Corvus monedula	R	1	0	0
Delichon urbica	LM	43	0	0
Dendrocopus major	PM	1	0	0
Emberiza citrinella	SM	23	0	0
Ficedula albicollis	LM	2	0	0
Ficedula hypoleuca	LM	31	0	0
Ficedula parva	LM	8	0	0
Hirundo rustica	LM	25	0	0
Jynx torquilla	LM	1	0	0
Locustella naevia	LM	1	0	0
Motacilla flava	LM	2	0	0
Muscicapa striata	LM	55	0	0
Passer montanus	R	37	0	0
Phylloscopus borealis	LM	1	0	0
Phylloscopus sibilatrix	LM	19	0	0
Picus viridis	R	1	0	0
Pyrrhula pyrrhula	PM	2	0	0
Saxicola rubetra	LM	1	0	0
Serinus serinus	SM	2	0	0
Sylvia borin	LM	42	0	0
Sylvia nisoria	LM	1	0	0
Turdus pilaris	SM	1	0	0
Turdus viscivorus	SM	1	0	0
Total number of specimens		4788	749	1339	–		56L, 27N	515L, 551N, 5U,	8L, 15N, 1A,	12L	1L, 2N, 1A	145

aLM: long-distance migrants (in many cases wintering in Africa); SM: short-distance migrants where the majority of individuals winter in Europe; PM: partial migrants where individuals either migrate short distances within Europe or remain resident; R: residents

bProbably Lanius collurio

cAbbreviations: L, larva; N, nymph; A, adult female; U, undetermined developmental stage due to lack of photos; I. r, Ixodes ricinus; I. f, Ixodes frontalis; H. p, Haemaphysalis punctata; H. m, Hyalomma marginatum; ND, not determined to genus, species or developmental stage because of lack of photos of ticks and/or loss of sample; SE, standard error; IQR, interquartile range

Fig. 1

Monthly patterns of birds infested with ticks during the study period. In brackets: number of bird captures infested by ticks/total number of bird captures checked for ticks during the collection period (Spring: 15 March–15 June; Autumn: 15 July–15 November 2009, shaded in gray). One bird capture was excluded from the analysis because of missing collection data

A total of 1339 ticks was collected from the infested birds. Due to missing photos of ticks and because some ticks were damaged when removed from their hosts, or lost in the nucleic acid extraction process, the genus, species, or developmental stage could not be determined in 10.8% (n = 145) of the ticks (Table 1).

Of the remaining ticks, for which the genus, species and/or developmental stage could be determined (n = 1194), 1.00% were Haemaphysalis punctata (n = 12; all larvae), 0.30% were Hyalomma marginatum (n = 4; 1 larva, 2 nymphs, 1 adult female), and 98.7% were Ixodes (n = 1178; 579 larvae, 593 nymphs, 1 adult female, and 5 ticks that could not be determined to developmental stage; Table 1). Among the Ixodes ticks, 93.0% (n = 1095) were identified to species of which 97.8% were I. ricinus (n = 1071; 515 larvae, 551 nymphs, 5 ticks that could not be determined to developmental stage), and 2.20% were I. frontalis (n = 24; 8 larvae, 15 nymphs, 1 adult female). Among all ticks of which the developmental stage could be determined, 49.8% were larvae (n = 592), 50.0% were nymphs (n = 595), and 0.20% were adult females (n = 2).

Larvae and nymphs of the most prevalent tick species, i.e. I. ricinus, were collected during the whole collection period but mainly from mid-March to mid-May and from early August to mid-October (Fig. 2a). Of all collected I. ricinus ticks, larvae were more prevalent in autumn (59.0%, 15 July–15 November) than in spring [36.%, 15 March–15 June, (χ2 = 53.93, df = 1, P < 0.0001]. I. frontalis and H. marginatum were only recorded on birds during the spring collection. H. punctata was recorded in both April–May and August (Fig. 2b).

Fig. 2

Temporal distribution of ticks removed from birds. Temporal distribution of a Ixodes ricinus and b all other tick species removed from birds during the collection period (spring: 15 March–15 June; autumn: 15 July–15 November 2009, shaded in gray)

Most ticks (46%) were found on the European robin (Erithacus rubecula), which was also infested with the highest number of tick species (n = 4) (Table 1). Among all ticks identified to species, I. ricinus infested the broadest range of bird species, i.e. 32 species (Table 1; Fig 3a). Most of the I. ricinus nymphs were removed from E. rubecula and Common blackbird (Turdus merula) (Table 1; Fig. 3b). Most I. ricinus larvae were removed from E. rubecula, Eurasian wren (Troglodytes troglodytes), and Tree pipit (Anthus trivialis) (Fig. 3c).

Fig. 3

Relative importance of different bird species as hosts for Ixodes ricinus. a All I. ricinus specimen, b I. ricinus nymphs, and c I. ricinus larvae. Bird species from which < 30 ticks were collected were included as “other bird species”

I. frontalis and H. punctata were each present on six species of birds (Table 1) and H. marginatum on three bird species, i.e. E. rubecula, Willow warbler (Phylloscopus trochilus), and Common chaffinch (Fringilla coelebs).

Prevalence of Borrelia bacteria and TBEV in the ticks

Of the ticks for which the genus, species, or developmental stage could be determined (n = 1194), 94.1% (n = 1124) were available for analyses of contents of Borrelia bacteria and TBEV.

Of all ticks analysed, 25.4% (285/1,124) were positive in real-time PCR for Borrelia spp. (Table 2). All 1124 ticks were negative in real-time PCR for TBEV. A significantly higher proportion of nymphs (36.2%, 206/569) than larvae (14.5%, 79/546 [χ2 = 69.18, df = 1, P < 0.0001]) was positive for Borrelia spp. Both adult tick females, i.e. one H. marginatum and one I. frontalis, were Borrelia spp. negative.

Table 2

Prevalence of Borrelia species in ticks removed from bird species captured at Ottenby Bird Observatory, Sweden 2009

Tick species	Stagea	No. examinedb ticks of each stage	No. (%) positive ticksc	No. (%) ticks containing Borrelia species determined by nucleotide sequencingd
				B. a	B. g		B. v		B. b		B. t		B. l	B. m			Mixed	UT
I. ricinus	L	514	80 (16)	1 (1)		19 (24)	3 (4)	1 (1)							8 (10)	2 (2)		46 (58)
	N	549	204 (37)	29 (14)		30 (15)	38 (19)	1 (1)		2 (1)		1 (1)			4 (2)	5 (2)		94 (46)
	U	5	1 (20)				1 (100)
I. frontalis	L	8	0 (0)
	N	15	3 (20)															3 (100)
	A	1	0 (0)
H. punctata	L	12	1 (8)															1 (100)
H. marginatum	L	1	0 (0)
	N	2	0 (0)
	A	1	0 (0)
Ixodes spp.	L	11	0 (0)
	N	5	1 (20)															1 (100)
All species	L	546	81 (15)	1 (1)		19 (23)	3 (4)	1 (1)							8 (10)	2 (3)		47 (58)
	N	571	208 (36)	29 (14)		30 (14)	38 (18)	1 (1)		2 (1)		1 (1)			4 (2)	5 (2)		98 (47)
	A	2	0 (0)
	U	5	1 (20)				1 (100)
All stages	L, N, A, U	1124	290 (26)	30 (10)		49 (17)	42 (14)	2 (1)		2 (1)		1 (1)			12 (4)	7 (2)		145 (50)

aTick developmental stage: L, larva; N, nymph; A, Adult female; U, undetermined

bTicks were examined for Borrelia spp. by the LUX real-time PCR assay

cNumber of ticks yielding a positive outcome by LUX real-time PCR assay

dAbbreviations of Borrelia species: B. a, Borrelia afzelii; B. g, Borrelia garinii; B. v, Borrelia valaisiana; B. b, Borrelia burgdorferi (s.s); B. t, Borrelia turdi; B. l, Borrelia lusitaniae; B. m, Borrelia miyamotoi; UT, untypeable

No significant difference was detected between the proportion of Borrelia spp.-positive ticks in real-time PCR collected in spring (24.6%, 15 March–15 June) and Borrelia spp.-positive ticks collected in autumn (26.5%, 15 July–15 November) (data not shown).

Among the I. frontalis ticks, 20.0% (3/15) of the nymphs were real-time PCR positive for Borrelia spp. (Table 2), while all larvae were negative (0/8). All 4 H. marginatum ticks were real-time PCR negative for Borrelia spp., whereas only 1 among 12 H. punctata larvae was positive for Borrelia spp.

Prevalence of Borrelia species in the ticks

Seven different Borrelia species were identified by sequence analysis of the 5S-23S IGS or 16S-23S IGS (Table 2). B. garinii was the predominant species and detected in 17% of all Borrelia-containing ticks, followed by B. valaisiana, 14%; B. afzelii, 10%; B. miyamotoi, 4.0%; B. burgdorferi s.s, 1.0%; B. lusitaniae, 1.0% and B. turdi, 1.0%, and a mix of different Borrelia spp., 2.0%. Based on the sequencing chromatograms, it was not possible to distinguish the specific Borrelia spp. involved in the mixed infections. Fifty percent of all Borrelia bacteria detected in the ticks were untypeable.

Of all samples that were determined to Borrelia species level (n = 138; Table 2), 52.9% (n = 73) were detected in ticks captured in spring (15 March–15 June), and 47.1% (n = 65) were detected in ticks captured in autumn (15 July–15 November) (data not shown). B. garinii was more prevalent in the autumn (60.0%, 39/65,) than in spring (20.5%, 15/73 [χ2 = 22.47, df = 1, P < 0.0001]). In contrast, B. valaisiana and B. miyamotoi were more prevalent in spring than in autumn (B. valaisiana: 45.2%, 33/73 vs. 13.8%, 9/65 [χ2 = 15.97, df = 1, P < 0.0001]; B. miyamotoi.: 13.7%, 10/73 vs. 3.1%, 2/65 [P = 0.03 (OR: 5 CI: 1.216-23.35)]). For the other Borrelia species there were no significant differences in seasonal prevalence.

Of all samples that were determined to Borrelia species level and where the developmental stage of the corresponding tick could be determined (n = 137), 23.4% (n = 32) Borrelia species were recorded in larvae while 76.6% (n = 105) were recorded in nymphs (Table 2). B. garinii was more prevalent in larvae (59.4%, 19/32) than in nymphs (28.6%, 30/105 [χ2 = 10.13, df = 1, P = 0.002]) (Fig. 4). B. miyamotoi was also more prevalent in larvae (25.0%, 8/32) than in nymphs (3.8%, 4/105 [P = 0.001 (OR: 8.417 CI: 2.565-26.25]). B. afzelii and B. valaisiana showed the opposite pattern and were more prevalent in nymphs (27.6%, 29/105, and 36.2%, 38/105, respectively) than in larvae (3.1%, 1/32 [P = 0.003 (OR: 11.89 CI: 1.990-125.2)], and 9.4%, 3/32 [P = 0.004 (OR: 5.483 CI: 1.717-17.9)], respectively). There were no significant differences in prevalence between nymphs and larvae among the other Borrelia species.

Fig. 4

Distribution of Borrelia species in larvae (L) and nymphs (N). The percentage of Borrelia spp. PCR-positive ticks per typeable species is given. The numbers of infected ticks examined are indicated above bars. **P < 0.01; ***P < 0.001. B. a, B. afzelii; B. g, B. garinii; B. v, B. valaisiana; B. b, B. burgdorferi (s.s); B. t, B. turdi; B. l, B. lusitaniae; B. m, B. miyamotoi

High proportions of I. ricinus larvae containing B. garinii were removed from Common whitethroat (Sylvia communis) and A. trivialis (Fig. 5a). Also, a high percentage of tick larvae containing B. miyamotoi was removed from E. rubecula (Fig. 5b).

Fig. 5

Relative importance of different bird species as carriers of B. garinii and B. miyamotoi containing larvae. The Borrelia species presented were detected in I. ricinus larvae removed from birds

Quantity of Borrelia cells in the ticks

According to the LUX real-time PCR assay, the number of Borrelia cells per tick containing a typeable Borrelia species ranged from 2.0 × 100 cells to 7.0 × 105 cells (Fig. 6), with a median of 1.1 × 103. No significant difference in number of Borrelia spp. cells between larvae and nymphs was observed (medians 2.5 × 103, and 1.1 × 103, respectively).

Fig. 6

Borrelia species plotted against the number of Borrelia spp. cells per tick. Horizontal lines indicate the median, with upper and lower quartiles. *P < 0.05; ***P < 0.001. B. a, B. afzelii; B. g, B. garinii; B. v, B. valaisiana; B. b, B. burgdorferi (s.s); B. t, B. turdi; B. l, B. lusitaniae; B. m, B. miyamotoi. Due to few samples, B. b, B. t, and B. l were excluded from the statistical analysis

Ticks containing B. miyamotoi had significantly higher numbers of Borrelia cells (median of 2.0 × 104 cells per tick) compared to ticks with B. afzelii (median 3.0 × 102, [Z = 4.175, df = 6, P < 0.001]) and compared to ticks with B. valaisiana (median 6.0 × 102, [Z = 3.824, df = 6, P < 0.001]) (Fig. 6). Ticks with B. garinii had significantly higher numbers of Borrelia cells (3.0 × 103 cells per tick) compared to ticks with B. afzelii (median 3.0 × 102, [Z = 2.644, df = 6, P = 0.049]). No other significant differences in the number of Borrelia cells among Borrelia species were detected. Ticks with untypeable Borrelia, had a significantly fewer Borrelia cells (median of 2.0 × 100) compared to ticks with a typeable Borrelia species (median of 1.1 × 103, P < 0.0001) (data not shown).

Discussion

Tick infestation prevalence on birds

In the present study, tick infestation rates were highest in late April and late September (Fig. 1), when as many as 30% of investigated passerine birds were infested with at least one tick. However, ticks were detected on birds in all study months, with infestation rates in the range of 4.90–31.5%. Observed infestation rates are likely influenced both by the phenology of migration of particular bird species and by tick and host biology. For instance, the early season (March–April) is dominated by short-distance migrants, including important ground foraging tick host species such T. merula and E. rubecula, while the latter part of spring (May–June) is dominated by long-distance tropical migrants, that to a larger extent comprise insectivorous species that feed less on the ground. In fact, in the present study, most ticks were removed from E. rubecula, T. merula, T. troglodytes, and A. trivialis (Table 1). These results conform to several similar studies, which have revealed that passerine bird species that frequently feed on the ground are more often infested with I. ricinus ticks [2–4, 12, 16, 32–35]. In a recent study in Denmark, E. rubecula, T. merula, and P. phoenicurus constituted less than one third of the birds examined for ticks, but carried 77% of all I. ricinus ticks collected; 78% of all 179 I. ricinus ticks were collected from birds in the autumn [4]. Similarly, in Switzerland, the prevalence of tick infestation, mainly of I. ricinus on migratory birds was significantly higher on birds migrating southwards than northwards [9]. In our study, the highest tick infestation prevalence occurred in April and August–September (Fig. 1).

Birds as reservoirs and transmission hosts for Borrelia species

A few xenodiagnostic investigations have demonstrated that birds are competent transmission hosts for certain Borrelia species. This suggests that birds may harbour infectious Borrelia spirochaetes in their skin or blood and consequently can act as competent reservoirs. Another possibility is that birds may be competent transmission hosts in non-spirochaetemic co-feeding transmission among one or more Borrelia-containing ticks, feeding adjacent to one or more susceptible ticks. A third possibility is that certain birds may act in both ways of Borrelia transmission. Kurtenbach et al. (2002) showed that pheasants are competent transmission hosts for B. garinii and B. valaisiana, but not for B. afzelii [36]. Comstedt et al. (2006) obtained evidence that several species of birds act as transmission hosts for B. garinii, thus transmitting the bacteria to larvae of I. ricinus, which were ingesting blood from their avian host [2].

The present study may suggest that many I. ricinus larvae contracted B. garinii and B. miyamotoi spirochaetes while feeding on some of their avian hosts (see below). However, while transovarial transmission is rare in B. burgdorferi (s.l.) [37], it is well known that such transmission is a common trait in B. miyamotoi, which results in B. miyamotoi-containing larvae capable of transmitting the spirochaetes to their hosts [37]. Thus, it is most likely that B. miyamotoi-containing larvae had been transovarially infected whereas the B. garinii-containing larvae had contracted the spirochaetes from their avian hosts.

Tick species

Ixodes ricinus

I. ricinus is considered to be the primary vector of all human-pathogenic species of B. burgdorferi (s.l.) and of B. miyamotoi occurring in Europe [38]. Moreover, I. ricinus is able to transmit B. turdi to T. merula [32]. Larvae and to a lesser extent nymphs of I. ricinus parasitize small mammals, birds, and reptiles; all stages can also be found on medium- and large-sized mammals [39, 40]. In most investigations, the number of larvae of I. ricinus collected from birds is usually larger than the number of nymphs collected. However, the proportion of nymphs of I. ricinus in relation to that of larvae of I. ricinus is usually higher on birds than on small mammals. For instance, during the summer months of 1991–1994 in a study area near Stockholm, Sweden, median numbers of 16–105 larvae and 0–2 nymphs of I. ricinus were found on infested bank voles, Myodes glareolus [40]. In contrast, from 22,998 birds captured at eight localities in Sweden, Olsen et al. [33] removed 949 I. ricinus ticks; 29.8% were larvae and 70.0% were nymphs. Even though the nymphs are easier to detect than the larvae (some of which may therefore have been missed), it is generally accepted that nymphs constitute a significantly higher proportion of the I. ricinus ticks on birds than on small mammals. This is confirmed by the results of the present study: almost equal proportions of larvae (48.3%) and nymphs (51.7%) of I. ricinus were recorded. Almost equal proportions of larvae and nymphs of I. ricinus were also encountered in similar studies by others [4, 41–43]. In the present study 46% of the ticks, most of which were immatures of I. ricinus, were removed from E. rubecula. Other bird species from which high total numbers of ticks were collected were T. merula, T. troglodytes, and A. trivialis. The large numbers of ticks from these bird species partly reflects their abundance, where E. rubecula, T. merula, and T. troglodytes (and to a lesser extent A. trivialis) make up a large proportion of the trapping totals, but it likely also reflects their habits of spending a lot of time on the ground, thereby rendering them vulnerable to attack by questing larvae and nymphs of I. ricinus.

Since the prevalence of tick-associated pathogens is generally higher in tick nymphs than in tick larvae, the high proportion of nymphs feeding on birds suggests that birds receive a substantial proportion of infections from the nymphs. On the other hand, the tick larvae, while feeding on their avian hosts, are likely to contract Borrelia bacteria with nymph-derived spirochaetes. In the present study, in I. ricinus larvae we detected these Borrelia species: B. afzelii, B. garinii, B. valaisiana, B. burgdorferi (s.s.), and B. miyamotoi. B. burgdorferi (s.l.) spirochaetes are usually not transmitted transovarially [37]. Therefore, we think it probable that the larvae, containing one or two species in the B. burgdorferi s.l. complex (Table 2), had contracted the Borrelia cells directly from their infective avian hosts or through co-feeding transmission. In contrast, B. miyamotoi is often transmitted transovarially. Therefore, we cannot know if the eight I. ricinus larvae positive for B. miyamotoi had contracted the spirochaetes transovarially or from their infective avian hosts or by co-feeding transmission.

Ixodes frontalis

Among the Ixodes ticks that were identified to species level, 2.0% were I. frontalis (8 larvae, 15 nymphs, and 1 adult female tick). Untypeable Borrelia spp. were detected in 3 among 15 I. frontalis nymphs.

I. frontalis parasitizes birds, particularly Passeriformes [44], and is a likely enzootic vector of blood-borne microbes in bird-tick-bird cycles. I. ricinus, which feeds on both birds and mammals, could here potentially function as a bridge vector by transferring pathogens from the bird-I. frontalis enzootic cycle to humans and other mammals. Gilot et al. (1997) describes one occasion when an adult female of I. frontalis attached to a human tick-collector’s hand [45]. This is apparently an exceptionally rare behaviour exhibited by this otherwise strictly ornithophagous tick species. In western and central France and in north-eastern Spain, I. frontalis is the most abundant ixodid on passerine birds [46, 47]. Heylen et al. (2016) have shown that I. frontalis is able to transmit B. turdi to T. merula, but that I. frontalis appears to be an incompetent vector of B. garinii and B. valaisiana [32].

Agoulon and co-workers showed that the seasonal activity of I. frontalis differs from that of I. ricinus, particularly of the larvae. In I. frontalis larval activity in France was completely absent during the summer [47]. This conforms to our results: immatures of I. ricinus were recorded during both spring–early summer and late summer–autumn while I. frontalis was only present during the first of these seasons.

Haemaphysalis punctata

The adults of Haemaphysalis punctata are usually parasites of medium and large mammals, in particular domestic ungulates, but may occasionally feed on humans [39, 44, 48, 49]. The immatures can be found on small mammals and more rarely on birds and lizards [39, 44, 48]. In the present study, we recorded 12 larvae of H. punctata, 11 of which were investigated and scored negative for Borrelia spp. (Table 1). Previously, among 967 ticks removed from 465 tick-infested birds out of nearly 23,000 birds examined, only 1 H. punctata, a nymph, was found [2]. An earlier study on questing ticks at three Swedish Baltic Sea islands, spirochaetes, believed to belong to B. burgdorferi (s.l.) were recorded in ~ 2% of H. punctata nymphs [50]. Since the spirochaetes were detected by phase-contrast microscopy, it is not certain that they belong to B. burgdorferi (s.l.). In any case, the low number of spirochaetes detected in the ticks and the low prevalence suggest that H. punctata is not an important Borrelia spp. vector [50]. This conclusion conforms to our results.

Hyalomma marginatum

Each spring Hyalomma marginatum is introduced to northern Europe when larvae or nymphs are carried by birds migrating from southern Europe to Sweden and neighbouring countries [39, 51]. This is reflected in our data by the fact that H. marginatum was only present on birds captured during April and May (Fig. 2b). During the summer of 2018, which was exceptionally warm, there were several records of adult H. marginatum and H. rufipes in Sweden [52]. These adult ticks are believed to originally have infested avian hosts in the Mediterranean region, then feeding and becoming nymphs while still on their hosts flying northwards for several days. Subsequently, the blood-fed nymphs left their hosts after their arrival in southern or central Sweden. Due to the warm summer of 2018 such H. marginatum and H. rufipes nymphs in Sweden were able to reach the adult stage before the end of the summer. However, these tick species are presumably not yet permanent members of the Swedish tick fauna. H. marginatum and H. rufipes are of great importance as primary vectors of CCHFV [53, 54]. In the present study, one larva, two nymphs, and one adult female of H. marginatum were recorded. All were negative for Borrelia. It is noteworthy that not only larvae but even nymphs and adult females can be transported into Sweden by birds. The fact that one adult female tick was encountered on a Fringilla coelebs suggests that, if/when the climate in southern Sweden becomes optimal for H. marginatum and H. rufipes, they will presumably establish indigenous populations here.

Borrelia garinii and Borrelia valaisiana

Birds are reservoirs and transmission hosts for human-pathogenic B. garinii and B. valaisiana [5, 11, 38, 55]. As shown in Table 2, most Borrelia-containing Ixodes larvae were positive for B. garinii. We believe that these larvae had contracted the B. garinii bacteria from their avian hosts. However, in a few cases where one or more Borrelia-containing nymphs were feeding together with larvae on the same bird, it cannot be excluded that the larvae contracted the spirochaetes by co-feeding transmission. The majority (69%) of the I. ricinus larvae with B. garinii were removed from either S. communis (42%) or A. trivialis (27%), which may indicate that these bird species may be competent reservoirs and/or transmission hosts for B. garinii. Among the species in the B. burgdorferi (s.l.) complex present in Europe, B. garinii has the greatest potential to cause severe late neurological disease manifestations. Therefore, from a public health point of view it can be considered to be the most important Borrelia species in Europe. This also underlines the view that birds have a considerable, although indirect, impact on the epidemiology of tick-borne human diseases. B. valaisiana has been found to be common in ticks removed from birds in Europe [2]. Surprisingly, only a few (n = 3) Borrelia-containing I. ricinus larvae removed from the birds in the present study proved to contain B. valaisiana spirochaetes. However, it should be noted that 46 larvae (58%) contained untypeable Borrelia bacteria. It is possible that some of them were due to B. valaisiana.

Borrelia afzelii

Borrelia afzelii was present in several nymphs of I. ricinus, but only in one larva. B. afzelii is generally considered to use small mammals as its vertebrate reservoir [38]. However, there are many reports describing the presence of B. afzelii in I. ricinus larvae removed from birds [38, 56, 57]. The reason(s) for the presence of B. afzelii in tick larvae attached to birds need(s) further investigations. The presence of B. afzelii in Ixodes nymphs removed from birds is most likely because these ticks, in their larval stage, had fed on B. afzelii-infected small mammals.

Borrelia lusitaniae

There are a few cases of human Lyme borreliosis where B. lusitaniae is considered the etiological agent. It has lizards of the Lacertidae family as its vertebrate reservoir and I. ricinus as its vector [58, 59]. B. lusitaniae is considered to have originated in Portugal from where it has spread to other countries in southern Europe and to North Africa and to scattered localities in Central, Eastern and South-Eastern Europe. DNA of this reptile-associated Borrelia species has even been demonstrated on a few occasions in adults and nymphs of I. ricinus that had fed on humans in southern and south-central Sweden [27]. The main method by which this Borrelia species has spread is thought to be by birds infested with B. lusitaniae-containing ticks. It has been recorded in Switzerland from I. ricinus larvae and nymphs feeding on E. rubecula and from larvae feeding on Turdus philomelos, T. merula and Phoenicurus phoenicurus [9].

Borrelia turdi

Birds are considered as reservoirs and transmission hosts for B. turdi. In Europe, I. frontalis is considered to be the primary vector of B. turdi [11, 32]. We did not record any among 16 I. frontalis specimens positive for B. turdi. However, 2 among 549 nymphs of I. ricinus were positive for this species. To our knowledge, B. turdi has not previously been recorded from ticks or birds captured in Sweden. However, B. turdi was first recorded in Europe by Hasle et al. (2011) in Norway [3], where they detected B. turdi in 0.4% of I. ricinus larvae and 0.4% of I. ricinus nymphs collected from northward-migrating passerine birds. B. turdi was then detected in I. frontalis, I. ricinus and H. punctata removed from birds in northern Spain [60], from I. ricinus removed from T. merula in Poland [5], from I. frontalis removed from T. merula captured in the Archipelago of the Azores [35], from I. frontalis feeding on birds in Belgium [61], and from I. frontalis feeding on T. merula, T. philomelos, P. major, and T. troglodytes in Portugal [62]. Prevalences of 7% to 56% of B. turdi in I. frontalis have been recorded [62]. B. turdi has also been detected in skin biopsies from T. merula, which is a competent transmission host for this bacterium [11, 62]. There is not yet any indication that B. turdi infects humans.

Borrelia burgdorferi (s.s.)

Only one I. ricinus larva and one I. ricinus nymph were positive for B. burgdorferi (s.s.). Several studies have shown that birds are competent reservoirs and transmission hosts for North American strains of B. burgdorferi (s.s.). [38]. B. burgdorferi (s.s.) is responsible for the majority of cases of Lyme borreliosis in North America. In Europe, B. burgdorferi (s.s.) is recorded rarely in ticks removed from birds [2, 5, 38, 57]. However, as pointed out by Franke et al. [56], this may just reflect that this species is generally rare in many European regions, rather than suggesting that birds are incompetent reservoirs for European strains of B. burgdorferi (s.s.).

Borrelia miyamotoi

Borrelia miyamotoi is transmitted transovarially from the adult female tick to her offspring, which results in B. miyamotoi-containing tick larvae [37]. Transovarial transmission is much rarer or absent in the species within the B. burgdorferi (s.l.) complex [37]. In the present study, eight I. ricinus larvae were positive for B. miyamotoi. Presence of spirochaetes in tick larvae infesting birds can occur if the avian host is infected with spirochaetes, if co-feeding transmission from a Borrelia-containing tick to an adjacent susceptible tick larva takes place or if transovarial transmission of spirochaetes from an adult female tick to her offspring occurs. Detection of spirochaetes in unfed Ixodes larvae is an indication of transovarial transmission, which is a prevalent trait in B. miyamotoi [37, 63]. This species is known to have rodents as one of its vertebrate reservoirs, but birds are considered to be another of its reservoirs [63, 64].

It is reasonable to assume that the infectiousness of a bacterial pathogen is, in general, positively correlated to the concentration of cells of the bacterium in the infective medium. Our results revealed a significantly higher median density of cells in B. miyamotoi-containing ticks than in B. burgdorferi (s.l.)-containing ticks. In another study, based on ticks infesting humans, Wilhelmsson et al. also recorded a higher median density of spirochaetes in B. miyamotoi-containing ticks than in B. burgdorferi (s.l.)-containing ticks [27]. We conjecture that this may be an evolutionary trait which compensates for the generally lower prevalence of B. miyamotoi in the tick population compared to that of B. burgdorferi (s.l.). Similarly, it has previously been shown that in rodents infected with either B. miyamotoi or B. burgdorferi, the concentration of cells in the blood was generally much higher in the B. miyamotoi-infected rodents [63]. It is important to realise that, since transovarial transmission is a common trait in B. miyamotoi, which results in B. miyamotoi-containing tick larvae, many of which will infest birds, this Borrelia can be dispersed to distant localities.

Tick-borne encephalitis virus in ticks infesting birds

Proven, competent vertebrate transmission hosts of TBEV are many species of small mammals (rodents, insectivores) while, in view of recorded viremia and/or virus isolations, some ungulates (goat, sheep) and many bird species [65, 66] may be considered suspected or presumed transmission-competent hosts for this virus. Moreover, there are a number of records of TBEV from ticks infesting birds [20, 34, 66, 67]. The prevalence of this virus in host-seeking I. ricinus and in ticks removed from hosts in TBEV-enzootic regions is usually < 1% [68]. The absence of TBEV in all bird-derived ticks investigated by us thus conforms to similar studies, which have recorded a low or zero prevalence of this virus.

Conclusions

The results corroborate the view that the contribution of birds to human Lyme borreliosis is substantial. Several ground-foraging bird species are directly and indirectly important for the local maintenance and long-distance dispersal of human-pathogenic or potentially pathogenic Borrelia bacteria. Our data suggest that S. communis and A. trivialis are reservoirs and/or transmission hosts for B. garinii. Many ground-foraging bird species are important tick hosts, which significantly contribute to the local maintenance of tick populations and to long-distance dispersal of several tick species.

Additional file 1: Fig. S1. The aligned Borrelia nucleotide sequences based on PCR products.