Unorthodox approaches can yield unexpected findings about host-vector-parasite interactions. In this case, we aimed to discover whether closely related raptor species attract different simuliid species which may transmit distinct parasite lineages. In the process, we used natural hosts as bait, combined with manual netting. This approach helped provide new knowledge about the diet of some of the most common European blackfly species. While most studies typically consider only engorged blackflies (but see [7, 14]), we succeeded to derive more information on diets from analyses of the presence of Leucocytozoon-lineages, combined with existent knowledge on their typical host groups.
Blackfly species and host preference
One of the main findings of our study was that certain blackfly species forage high up in the canopy and appear to be attracted to avian hosts. All but two out of the 154 identified individuals belonged to the subgenera Nevermannia and Eusimulium, and nearly 30% of the individuals of both subgenera had fed on birds, as revealed by their Leucocytozoon load. This finding indicates that Nevermannia and Eusimulium are the dominant ornithophilic subgenera of the genus Simulium and as such being the most probable vectors of Leucocytozoon-lineages in central Europe (c.f. ). This corresponds to findings that other European species of Nevermannia attack thrushes and warblers, while blackflies of the S. (N.) vernum group and S. (E.) aureum group are relatively rare at heights under 10 m in spruce and pine forests [9, 13, 28]. Before this study, feeding preferences of Nevermannia and Eusimulium species were known only from S. (N.) silvestre, S. (N.) curvans, S. (E.) angustipes and S. (E.) aureum, the latter being one of the best examined ornithophilic blackflies and vectors of Leucocytozoon [7, 11, 14, 15, 28, 29]. This species, however, appears to be rather rare in the upper canopy.
The composition of Leucocytozoon lineages found in this study shows that Nevermannia and Eusimulium species attack most avian host groups occurring in this canopy layer, which are sufficiently large and/or abundant, such as thrushes, corvids, pigeons, raptors, owls and tits. A similar size-and-abundance pattern of prey was found among ornithophilic and mammalophilic blackflies close to ground level in Scandinavia . In contrast to the pattern found in Scandinavia, we did not find a strong association between blackfly species or haplotype and Leucocytozoon lineages, or their corresponding vertebrate host group . In the study by Hellgren et al. , a limited number of engorged S. (N.) silvestre suggested a preference for thrushes. However, a much larger sample of S. (N.) silvestre from North America harboured Leucocytozoon lineages infecting avian species across the phylogeny, inferring that this species has a broad range of bird species in its diet . This pattern corresponds much better to our findings from central Europe and supports the notion that species of Nevermannia and Eusimulium have habitat preferences but are otherwise indiscriminately ornithophilic (Fig. 1, Table 1). Our results deliver no information in which vector species the respective Leucocytozoon lineage can complete their development and life-cycle. However, experimental evidence suggests that most parasites of the genus Leucocytozoon are more restricted by the ecology of the vector than by its physiology .
The absence of non-Leucocytozoon parasites also provides insight into the behaviour of blackflies. Plasmodium and Haemoproteus are not transmitted and cannot fulfil their development in blackflies, but our protocol was apt to detect them, and their prevalence in the putative hosts is relatively high [15, 30]. Therefore, their complete absence confirms that blackflies are not active in the foraging habitat after feeding and remain distant and inactive until the blood meal and potential abortive stages of Plasmodium and Haemoproteus are digested . Individual blackflies possibly return after oviposition, as the period between two feedings has been measured to take 5–7 days in S. rugglesi .
To our knowledge, this is the first study of haemosporidian parasites being present in individual non-engorged blackflies. We found that nearly 30% of the blackfly individuals active in the upper canopy layers of central Europe are Leucocytozoon-carriers. Blackflies are likely to return to a large stationary food source such as a raptor brood after oviposition [31, 32]. Nonetheless, 30% infected vectors are likely representative of the blackfly population in this habitat, since the raptor lineages potentially belonging to our “bait” accounted for only 15% of all infected blackflies. Previous studies have either analysed pools of non-engorged blackflies or individual visibly engorged blackflies [11, 13, 14, 29]. Engorged blackflies are not active after feeding (0.1–0.3% of all blackfly individuals caught in forests), but can account for up to 23.6% of all blackflies caught with a sweeping net in an alpine habitat [28, 33]. The frequency of Leucocytozoon-carriers among freshly engorged blackflies in Scandinavia was 62%, being more representative of the Leucocytozoon prevalence in avian hosts, which is expected to be higher there than in central Europe [13, 15]. On the other hand, close to 46% of pools of five non-engorged blackflies seem to contain Leucocytozoon lineages, suggesting a Leucocytozoon prevalence of approximately 20% in the corresponding blackfly populations [11, 14, 29]. Thus, we complement previous studies of Leucocytozoon-carrying blackflies with an individual-based estimate, which may be more precise but is specific to the upper canopy habitat of central Europe.
Vector and host habitat choice
Finally, we found a substantially different composition of blackfly species around the nests of three closely related avian hosts. Simulium (E.) rubzovianum was overrepresented around nests of red kites, and S. (N.) vernum* was the only blackfly species present around goshawk nests. At the same time, all species and the greatest diversity were represented around nests of common buzzards. This pattern could be due to host preference. Nests of the three raptor species can potentially be identifiable by odours, which is a primary sense for prey recognition outside of the visible range of blackflies [4, 34]. Buzzards for example, commonly have dead and decaying voles deposited around the nest. Red kites incorporate a great share of carrion and garbage in their nests and food, which lead to a distinct smell of the whole brood. Goshawks, on the other hand, feed mainly on birds and do not keep unconsumed prey remains at the nest. Although blackflies are not attracted to carrion per se, these compounds in addition to the native bird odours may enhance distinction of raptor species. However, it seems unlikely that the involved blackfly species discriminate against any of the raptor species, given the patterns outlined by the distribution of Leucocytozoon lineages found in this study.
Alternatively, the choice of breeding habitat by the three raptor species may predispose them to a different exposure of blackfly species around their nests. Red kites have a preference for open, dispersed deciduous and mixed forests, while goshawks prefer the core of bigger forests with a higher proportion of coniferous trees, and buzzards cover the whole continuum from single trees to the core of big forests. These preferences may co-vary with the microhabitat foraging preferences of the different blackfly species, which are very poorly known . Such a difference in the blackfly community, however, may have catalysed an ecological speciation of parasites, leading to the cryptic Leucocytozoon species infecting currently sympatric raptor species [16, 17, 35]. The substantial difference in sample size between the nests of the three raptor species does not allow us to adequately compare Leucocytozoon-lineage diversity around those nests and should be compensated by future studies.