For this study, we considered ticks to be sentinels of CCHFV distribution in Spain. Questing or feeding ticks were collected in 2 different phases from wild and domestic animals across a wide geographic area of Spain during October 2016–2018 (Figure 1). In the first phase, which lasted 6 months (October 2016–March 2017), ticks were collected from livestock flocks in 11 counties in 4 regions, CyL, Madrid (MAD), Castile-La Mancha (CLM), and EXT, where Hyalomma ticks were already known to be present (Figure 1). Earlier, smaller surveys had established the presence of CCHFV in both ticks and humans (11,18,20). Adult ticks were collected mainly from wild large game ungulate species such as red deer, wild boar, fallow deer (Dama dama), and mouflon (Ovis orientalis musimon) and barbary sheep (Ammotragus lervia), as well as from grazing livestock cattle (Bos primigenius taurus) and goats (Capra aegagrus hircus) that had not recently received an ixodicidal treatment and had been grazing on farms for ≥30 days. A maximum of 25 ticks per animal were collected; all ticks were removed from animals on which <10 ticks were found. In small herds of wild ungulates, all of the animals were sampled, but only 10%–15% of the members of larger herds were sampled.

Ticks from the same animal were kept alive in separate, labeled sterile vials and transported at ambient temperature with controlled humidity until they could be frozen at –80°C when possible, then sent to the Spanish National Centre of Microbiology (NCM) for tick identification and viral molecular analysis. Ticks were identified using taxonomic keys (22,23), and pools were produced with ≤4 fed ticks (depending on size and food status) with specimens collected from the same animal and of the same tick species. Ticks collected in CLM were identified at local laboratories and then sent to NCM. Ticks of the same species collected from different animals were combined in 1 tube.

In the second phase, during May–September 2017 and March–July 2018, ticks were collected from vegetation by standard flagging in 15 counties in 4 regions where the circulation of the virus had not previously been described: MAD, CLM, Murcia (MUR), and Andalusia (AND) (Figure 1). After ticks were collected and transported to the laboratory, they were morphologically identified (22,23), then samples were frozen and sent to NCM for molecular processing. Pools were produced with 3 ticks obtained from the field.

A total of 12,584 ticks were collected and pooled, 3,959 pools (10,793 ticks) from animals (Tables 1, 2, 3) and 597 pools (1,791 ticks) from vegetation (Table 4). Adult H. lusitanicum were predominant among ticks collected while feeding, but we also recorded H. excavatum, H. marginatum, H. rufipes, Dermacentor marginatus, Ixodes ricinus, Haemaphysalis punctata, Rhipicephalus annulatus, R. bursa, R. pusillus, and R. sanguineus sensu lato ticks (Table 1). We identified all but 2 questing ticks as H. lusitanicum.

Feeding Ticks

We collected ticks from 1,186 ungulates in CyL, MAD, and EXT: 943 wild and 243 domestic animals (Table 2). Red deer (n = 731) was the most common host species, followed by wild boar (n = 176), fallow deer (n = 19), and mouflon (n = 15) and barbary sheep (n = 2). Among livestock, we surveyed cattle (n = 235) more often than goats (n = 8). No ticks feeding on livestock were positive for CCHFV. Excluding animals from the CLM region, for which data were not available, we found positive ticks on 2.9% (28/943) of all surveyed wild animals.

Figure 2

Figure 2. Phylogenetic tree obtained for strains of Crimean-Congo hemorrhagic fever virus detected in Spain (black dots) and other sequences downloaded from GenBank. We built the phylogenetic tree using the neighbor-joining method...

We found CCHFV-positive feeding ticks in 128 (3.2%) of 3,959 pools (Table 3). We can therefore confirm that feeding ticks carried CCHFV RNA in 7 of 11 counties in 4 provinces in 4 regions: Cáceres (EXT), Madrid (MAD), Toledo (CLM), and Ávila (CyL) (Figure 1). We found marked differences in percentages of positive pools among regions: 1.5% in EXT, 0.3% in CyL, 11.8% in CLM, and 7.7% in MAD (Table 3). Ticks in which we detected CCHFV were mainly H. lusitanicum, although the virus was also found in I. ricinus (2 pools obtained from red deer), R. annulatus (1 pool from a fallow deer), and D. marginatus (1 pool from a wild boar) ticks. This finding does not indicate the vectorial characteristics of those tick species, because we collected them while they were feeding, but does indicate the breadth of distribution of the virus. With respect to the genetic variability of S segment sequences, in EXT we found genotypes I (Africa 1 clade), III (Africa 3 clade), and IV (Africa 4 clade) (Table 3; Figure 2); in MAD and CyL, we found genotype IV (Africa 4 clade), and in CLM, we found genotype III (Africa 3 clade).

After the diagnosis of the first human cases of CCHF in Spain in 2016 (11), a large field study was planned and implemented, with the support of human and animal health authorities, that aimed to estimate the geographic distribution of CCHFV in Spain. The virus is considered a serious human health issue, so risk assessment had to be based on a geographically wide-ranging campaign of viral detection, examining ticks collected while questing or feeding on diverse vertebrates. By testing for CCHFV RNA, we used ticks as sentinels for the presence of the virus. Although the study was not designed to estimate prevalence, our data showed a 2.96% rate of positivity (135 positive pools out of 4,556), close to the values reported from other endemic locations, such as Turkey (3.6%), Albania (3.2%), and Kosovo (3.6%) (27–29). However, the wide range of methods used to collect ticks and analyze samples make reliable comparisons difficult.

In total, 3,959 pools were processed from 10,793 collected ticks. Tick management with acaricide is routinely practiced with domestic animals so they rarely have ticks; therefore, most of the feeding ticks were collected from wild animals. All of the positive tick pools were obtained from wild ungulates, consistent with the higher CCHFV antibody prevalence found in wildlife (61%) compared with livestock (15%) in the same areas (30). Wild ungulates range freely in most of the studied territory, although some are farmed as game, where the animals are confined to large farms. On these farms, game animals can be exposed to immature Hyalomma ticks through close contact with prominent hosts, such as rabbits and hares, simultaneously explaining the greater abundance of infected ticks and the higher serologic titers among farmed wild ungulates than among livestock.

The significance of finding CCHFV in feeding ticks is always difficult to interpret, because the virus could have been acquired through the blood meal, which means that determining the ticks’ CCHFV status cannot provide an accurate estimate of actual infection rates. We used the feeding ticks as sentinels of the presence of CCHFV and as a means of comparing the prevalence of the virus noted in questing ticks. Although collecting questing ticks is always more expensive and time-consuming, combining data from both sources could help provide a more balanced perspective for large geographic areas such as entire countries.

We found CCHFV from I. ricinus, Rhipicephalus spp., and D. marginatus ticks in 4 pools, but we considered these data to reflect serendipitous detection of viral RNA in feeding ticks. All of our results point toward a predominant role for H. lusitanicum in CCHFV circulation. In fact, it is very likely that circulation of the virus is restricted to H. lusitanicum ticks, a factor that had been suspected (20,21) but not confirmed. This finding emerged from the second phase of the study, in which we surveyed questing ticks. We demonstrated that the virus can successfully complete cycles in nature while perpetuating itself in H. lusitanicum ticks. This finding adds an extra dimension to our results, because the range of H. lusitanicum ticks is rapidly expanding as a consequence of the spread of one of its natural hosts, wild boars. Although the vectorial status of this tick has not been demonstrated, the discovery of viral RNA in molted ticks after a blood meal demonstrates that the virus can at least persist in these ticks (5). It is therefore a matter of urgency to establish the vectorial status of the species, as well as its preferences for biting humans, which seems to be a promising field of research. Because this tick is a potential source of CCHFV infection, its ability to be a parasite in humans has clearly been neglected, probably because it has not been reliably identified in the few samples collected from humans.

We identified 5 regions in central and southwest Spain where CCHFV is present. Taken together, findings from our report and an earlier study (21) indicate the permanent circulation of CCHFV in these regions, which are characterized by a Mediterranean forest ecosystem rich in wild ungulates that undoubtedly favors the presence of H. lusitanicum ticks. The lack of positive detection in MUR, one of the surveyed regions, does not prove that the virus is not present there, but more likely reflects the small number of ticks collected from the region.

Our system of confirmation involved amplification by 2 methods or from 2 extractions amplified using the same method. We confirmed positive amplification by real-time or nested RT-PCR in 118 out of 135 pools. We could not confirm amplification in the other 17 pools and had to repeat extraction and real-time RT-PCR. Given that sensitivity in the 2 PCR methods is similar, these differences in amplification could be because of variability in the target region (31). We obtained sequences from 105 of 128 pools of ticks collected from animals and for 3 of 7 pools collected from vegetation, which enabled us to identify genotypes on the basis of their S segment sequences. Viruses belonging to Africa 3 (genotype III) and Europe 1 (genotype V) had previously been detected in ticks; the newly proposed Africa 4 clade should be added to these findings (18–21) and should be interpreted as arising through the continuous exchange of infected ticks by birds migrating between Africa and southern Europe. This survey also detected the circulation of Africa 1 (genotype I) in Spain. These results confirm the wider-than-expected distribution and broad variability of CCHFV in Spain. These findings were initially unexpected but are compatible with reports of the genetic variability of the virus, because CCHFV is well known to undergo rearrangement to produce diverse combinations of the S, L, and M segments (32).

The relationship of genotypes III, IV, and V to human cases in Spain has previously been described (Africa 3 in cases from 2016 and Africa 4 and Europe 1 from cases in 2018) (13,14,17). The great variability of genotypes may have resulted from multiple introduction events in Spain, but the complete mechanism of spread is very poorly understood. Published syntheses have reported that immature H. lusitanicum ticks feed on small mammals, perhaps mainly leporids, and adults feed on large ungulates (6). However, previous studies (33–36) have documented that when there are huge populations of the tick, large numbers of immature ticks were found on birds, such as the red partridge (Alectoris rufa), that spend most of their time on the ground.

Parasitism of birds does not seem to be the rule for H. lusitanicum ticks, which have not been found on birds migrating from Africa to Europe. We consider birds to be secondary, or even accidental, hosts for immature H. lusitanicum ticks, and therefore that immature H. marginatum ticks, which commonly feed on birds, may be the keystone species for transporting and importing CCHFV. We propose that annual migratory journeys from Africa of birds carrying H. marginatum ticks may have been the primary source of entry for several viral variants into Spain. Once introduced, the virus could have easily adapted to a cycle of transmission between wild ungulates and H. lusitanicum ticks, probably acquiring new mutations or reassortments. This hypothesis will be difficult to prove unless more data about H. marginatum ticks transported from Africa become available. Furthermore, the presence of European viral genotypes is difficult to reconcile with our observations and contrasts with data from other countries, such as Turkey and those of the Balkan region, where the virus is endemic. The occurrence of only 1 introduction from Asia to these countries has been proposed, and the strains causing human cases there have remained genetically stable for decades (37).

Our findings on the distribution of CCHFV in Spain demonstrate its presence in 5 regions covering the central and southwest part of the country. Our study also drew attention to the importance of H. lusitanicum ticks in circulating the virus including several viral genotypes and possible new reassortments. The risk for transmission to humans has not yet been possible to calculate because of the paucity of data. Research is needed to determine the reasons behind the high variability of CCHFV and the actual distribution and origin of circulating strains.

Clinicians, especially general practitioners, as well as laboratory staff, public health workers, stakeholders, and the general public need to be aware of the situation regarding CCHFV in Spain. Because some clinical cases may be mild and etiologically unresolved by practitioners, suitable tools must be made available that can detect the virus in suspected clinical cases in Spain. Diagnosis of CCHF is hampered by the biosafety conditions required to manage a virus of high biologic risk.

Public health activities, including surveillance of zoonoses like CCHF, need to be carried out under the One Health umbrella, as was done in our study. Large-scale seroprevalence studies in animals and humans are currently underway. The huge effort required to coordinate local and national public health representatives and entomologists, virologists, and animal and human health specialists should be an essential step in the control of these pathogens.