Breeding Habitat Preferences of Major Culicoides Species (Diptera: Ceratopogonidae) in Germany

Biting midges of the genus Culicoides (Diptera, Ceratopognidae) are demonstrably or putatively involved in the transmission of both bluetongue (BTV) and Schmallenberg viruses (SBV) in Central Europe. Although these insects are ubiquitous in Europe, relatively little is known about their requirements in terms of breeding habitats and substrates. Culicoides species composition and relative abundance in potential breeding habitats were therefore studied at various locations in Northeastern Germany and one location in Western Germany by emergence trap collections. Forty-three potential breeding sites were analyzed in ten landscape structures, with 28,091 adult biting midges emerging from them. Among these, 2116 specimens belonged to the genus Culicoides. Species of the culicoid subgenus Avaritia were most abundant (70.6% of all specimens) and widespread (91.6% of all sites), while the subgenus Culicoides accounted for 15.6% of the specimens registered but emerged from 70.8% of all sites sampled. Culicoides species of other subgenera were collected in 75.0% of all studied sites, with a relative abundance of 8.7%. The results indicate that various types of dung, but probably also some landscape habitats, offer suitable substrates for the development of potential Culicoides vector species. Adaptations in dung management on farms and landscape design and use might therefore be appropriate approaches to reduce the risk of BTV or SBV transmission.


Introduction
Due to their painful bites and often high abundances, female biting midges of numerous species of the genus Culicoides are well-known pests affecting humans and livestock [1]. In addition, some species are vectors of disease agents [2]. In Central and Northern Europe, the transmission of pathogens by biting midges was regarded as exotic until the advent of bluetongue disease (BT) in 2006 [3]. Globalization and modern transportation technologies involving long-distance animal movement facilitate the travel of both infected vertebrates and vectors to reach almost any place in the world within a short time [4]. Thus, when animals infected with a vector-borne disease agent are imported, indigenous competent vectors, which may ingest and replicate the pathogen, pose a significant risk of local transmission. In the case of small insects like biting midges, passive drift by wind over hundreds of kilometres is discussed, too, allowing for biting midge-borne pathogens to be introduced via their vectors [5].

Study Sites and Trapping Procedure
Freshly emerged adult biting midges were collected in Western and Northeastern Germany by emergence traps, from May to December 1994December , 2005December and 2007December -2012 (Table S1). The traps were located in ten wet or humid landscape structures, such as river banks and floodplains, along the rivers Oder, Spree and Rhin, in the reed belts of lakes, wetlands and grasslands (both with and without bushes and trees), swampy forests and renaturation areas. The latter were only represented by floodplains along small streams. These areas had once been converted to forests for wood production and to greenlands for silage production and have only recently been restored to their natural status as floodplains. Further traps were positioned in specific "habitats": on liquid manure, compost and different types of dung (two to three days old), on animal holdings and on meadows and pastures with animal husbandry in Oderbruch floodplain areas in the German federal states of Brandenburg and Saxony-Anhalt ( Figure 1). All collection sites had a damp underground.

Study Sites and Trapping Procedure
Freshly emerged adult biting midges were collected in Western and Northeastern Germany by emergence traps, from May to December 1994December , 2005December and 2007December -2012 (Table S1). The traps were located in ten wet or humid landscape structures, such as river banks and floodplains, along the rivers Oder, Spree and Rhin, in the reed belts of lakes, wetlands and grasslands (both with and without bushes and trees), swampy forests and renaturation areas. The latter were only represented by floodplains along small streams. These areas had once been converted to forests for wood production and to greenlands for silage production and have only recently been restored to their natural status as floodplains. Further traps were positioned in specific "habitats": on liquid manure, compost and different types of dung (two to three days old), on animal holdings and on meadows and pastures with animal husbandry in Oderbruch floodplain areas in the German federal states of Brandenburg and Saxony-Anhalt ( Figure 1). All collection sites had a damp underground. Geographical allocation of study regions sampled for emerging adult biting midges. A: Two locations with seven study sites; B: Two locations with three study sites; C: Two locations with three study sites; D: Four locations with six study sites; E: Two locations with four study sites; F: One location with two study sites; G. One location with two study sites; H: One location with six study Figure 1. Geographical allocation of study regions sampled for emerging adult biting midges. A: Two locations with seven study sites; B: Two locations with three study sites; C: Two locations with three study sites; D: Four locations with six study sites; E: Two locations with four study sites; F: One location with two study sites; G. One location with two study sites; H: One location with six study sites; I: Five locations with seven study sites; J: One location with two study sites; K: One location with one study site.
In addition to "landscape structure" and "habitat", the terms "(study) region", "location" and "(study) site" are used in this contribution. The "study region" refers to the greater geographical region where the study was performed, as depicted in Figure 1. A "study region" can include several "locations", which are smaller administrative units specified by the local name. Finally, a "study site" defines the exact place where an emergence trap was operated.
A detailed compilation of the geographic regions, locations, study sites, numbers of used traps and landscape structures, as based on the CORINE (Coordination of Information on the Environment) land cover index but modified for small-scale assessments, is provided in Table S1. CORINE land cover indices were used to relate land cover and use to Culicoides occurrence.
Traps of 1 × 1 m 2 ground area were mostly used, except for some dung and liquid manure substrates, where traps of 0.5 × 0.5 m 2 were used, based on the habitat dimensions. All data collected were converted to the 0.5 × 0.5 m 2 scale to facilitate comparison between collection sites.
Insects emerging from the soil covered by the traps were collected in plastic containers mounted on top of the traps and filled with 70% ethanol to kill and preserve the material. All traps were run once per month for 48 h.
At site 29 (River Oder, Güstebieser Loose), in the floodplains of the river Oder, site 31 (River Spree, Mönchwinkel), in the floodplains of the river Spree, and site 34 (Backwater River Oder, Platkow), on the banks of a backwater of the river Oder, emergence traps were set up on meadows where cattle, sheep or domestic ducks were maintained. For comparison, emergence traps were placed at three sites (sites 1, 2, 3) in the same region on floodplains without livestock and contamination by excreta. Like other floodplains of lowland rivers sampled in the study (sites 5, 7), these regions are characterized by open grasslands and dominated by abandoned agricultural meadows.
In addition to these field areas, eleven farms in the same region were included, according to the keeping of livestock producing various types of dung (Table S1). For setting the traps, both homogeneous and mixed types of dung were selected. As for the latter, mixed sheep/horse and goat/horse dung could be studied in Kiehnwerder (site 32) and mixed cattle/horse dung in Rosenthal (site 33).
Soil pH values were measured under each emergence trap using a Multi 3410 measuring device (WTW, Germany), and they were analyzed using SPSS 24.0 statistical software (IBM, Amonk, NY, USA).

Biting Midge Identification
Culicoides biting midges were sorted from the catches and morphologically identified mostly to species level or to a lower taxonomic resolution when this was not possible, using various Diptera-specific identification keys [36][37][38][39][40].
Within the subgenus Avaritia, the females of the Obsoletus Complex (C. obsoletus (Meigen, 1818), C. scoticus (Downes and Kettle, 1952)) cannot be morphologically distinguished. Additionally, the identification of C. chiopterus (Meigen, 1830) and C. dewulfi (Goetghebuer, 1935) is challenging, particularly so when caught by emergence traps. Firstly, specimens often become conserved in the fixative before hardening of the chitinous parts and acquiring their final coloration. Secondly, samples exposed to the sunlight, as was the case in the present study due to the transparent construction material of the collection boxes of the emergence traps used, tend to bleach. Therefore, only a fraction of the Avaritia females could be identified to species level. Thus, the focus was on males, as these can be reliably identified to species by their genitalia. If males were not available, randomly selected females were processed by species-specific diagnostic PCR assays [41,42]. Some culicoids were subjected to COI barcoding using primers LCO1490 and HCO2198 (Eurofins Genomics, Ebersberg, Germany) [43] or primers PanCuli-COX1-211F and PanCuli-COX1-727R (Eurofins) [42] for reliable species identification.

Data Analysis
Data were analyzed in R version 3.5.2 (R Foundation, Vienna, Austria) [44] using the vegan package [45]. Non-metric multidimensional scaling (NMDS) was used to plot the relative habitat preferences of the subgenera Avaritia and Culicoides and of the group of species belonging to other subgenera, as well as of species of subgenera other than Avaritia. The subgenus Avaritia was not considered in the second analysis, since only a fraction of the females was identified to species level and unidentified females could have a major impact on the data analysis should they not display the same species ratio as that of the identified specimens. To harmonize the collection data, the analysis was based on the average number of emergent culicoid specimens per trap and year. The suitability of the data was demonstrated by a stress test before applying the NMDS, using the Bray-Curtis similarity index with 100 iterations for reaching a stable solution.
In single 48-h catches, the percentage of Ceratopogonidae varied from 0.1% to 32%. Many ceratopogonid specimens belonged to the genus Forcipomyia and to further genera other than Culicoides (Tables 1 and 2). Culicoids were found in 817 collections (37.8%).
Considering individual landscapes and habitats, the total number of emerged culicoids varied between very few or even single specimens per 48 h in swampy forest areas (sites 16,17) or liquid manure habitats (sites 42, 43) to more than 50 in renaturation areas (site 6) and dung sites on farms (sites 32, 33).
Twenty-one species of the genus Culicoides were identified. In total, the proportion of biting midges of this genus accounted for 0.2% (2116 specimens) of the total insect catches. These included all four species of the subgenus Avaritia occurring in Germany (C. obsoletus, C. scoticus, C. chiopterus, C. dewulfi). Furthermore, four members of the subgenus Culicoides were found: C. delta (Edwards, 1939), C. pulicaris (Linnaeus, 1758), C. punctatus (Meigen, 1804) and C. newsteadi (Austen, 1921), with the latter three grouped into the Pulicaris Complex. The remaining Culicoides species belonged to subgenera other than Avaritia or Culicoides (Table 1).
Based on the male Avaritia and Culicoides specimens (female specimens were not included due to the Avaritia females not being reliably identified to species level), 92.6% of Culicoides specimens belonged to the Avaritia subgenus and 6.7% to the Culicoides subgenus. Culicoides obsoletus and C. scoticus as members of the Obsoletus Complex were particularly abundant (83.3% and 10.4% of all Avaritia males, respectively), while C. chiopterus (5.2%) and C. dewulfi (1.1%) males were found in low numbers. In the subgenus Culicoides, C. delta accounted for most of the collected specimens (37.1% including females).
Total number of females/males of other subgenera 69/0 In "mixed grassland/wetland with bushes and trees", subgenus Culicoides specimens accounted for 33.4%, while the species of the subgenus Avaritia, which was represented by the Obsoletus Complex only, accounted for 41.9%, and species of other subgenera (C. achrayi, C. fagineus, C. minutissimus, C. nubeculosus) for 24.5% of the culicoid biting midges. Banks of rivers and lakes seemed to be attractive for members of the subgenus Avaritia (C. obsoletus, C. scoticus, C. chiopterus), which made up 47.6%, while individuals of other subgenera (C. achrayi, C. circumscriptus, C. fagineus, C. kibunensis) reached 41.8% and those of the subgenus Culicoides (C. pulicaris, C. punctatus) reached 10.4% of all culicoid biting midges collected. Single individuals of three species of other subgenera, C. nubeculosus, C. riethi and C. festivipennis, were found in liquid manure.
According to the NMDS analysis, which calculates a relative habitat preference (i.e., the result for one subgenus/species depends on the consideration of all other subgenera/species), the subgenus Avaritia was strongly associated with compost and dung habitats and the subgenus Culicoides with wetland, grassland, floodplain, meadows, mixed grassland/wetland with and without bushes and trees and reed belt landscapes, and the group of species belonging to other subgenera was associated with swampy forests and the banks of rivers and lakes (Figure 2A). In this analysis, it should be taken into account that the grouping eliminates ecological differences between species, which is particularly relevant with regard to the species belonging to the subgenera Avaritia and Culicoides as these are systematically more mixed and might be only distantly related.
When individual sites are considered, the highest numbers of Culicoides species and specimens emerged from dung sites (sites 24, 29, 33, 36) and a renaturation area/floodplain (site 6) and the lowest number from meadows (site 5). In the subgenus Avaritia, high numbers of specimens were obtained from dung habitats (sites 32, 33, 36), floodplains (site 1) and compost (sites 24, 26), and low numbers were obtained from swampy forests (sites 16, 17) and boglands (site 23). The highest numbers of subgenus Culicoides specimens were collected in renaturation areas (site 6), mixed grassland/wetland with bushes and trees (sites 9, 11) and reed belts (site 21), and low numbers were collected from swampy forests (site 16) and meadows (site 4).
The highest number of a species of the subgenus Avaritia, C. obsoletus, was collected from cowpats and llama/alpaca dung (sites 29, 36), while in the subgenus Culicoides, C. delta reached the highest collection numbers in a renaturation area (site 6). Among the species belonging to other subgenera, C. achrayi was the most abundant, namely in mixed grassland/wetland with and without bushes and trees (sites 12, 14, 15) and on the banks of rivers (site 21).
The NMDS analysis for landscape/habitat preferences of species (except for Avaritia species) shows C. sphagnumensis to be most closely associated with reed belts; C. manchuriensis and C. circumscriptus with floodplains; C. kibunensis with the banks of rivers and lakes; C. minutissimus with mixed grasslands/wetlands without bushes and trees and the banks of rivers and lakes; C. achrayi with mixed grassland/wetland without bushes and trees; C. punctatus with mixed grassland/wetland without bushes and trees and boglands; C. vexans and C. impunctatus with boglands; C. pulicaris with boglands and reed belts; C. nubeculosus with mixed grassland/wetland without bushes and trees and renaturation areas; C. delta with renaturation areas; C. riethi with liquid manure; C. albicans, C. fagineus and C. festivipennis with swampy forests and C. newsteadi with meadows ( Figure 2B).
Special attention was given to the study of dung as a potential breeding substrate, as species of the subgenus Avaritia are supposed to preferentially develop in organic-rich material [46]. Accordingly, pure dung substrates, where Avaritia species in fact predominated, produced twice as many specimens as the studied landscape habitats, with a particularly clear preference of both species of the Obsoletus Complex, C. obsoletus and C. scoticus, as well as C. chiopterus for dung habitats (Table 1). Depending on the origin of the dung, differences in the abundance and composition of the various emerging species became apparent (Table 2).
Culicoides obsoletus was mainly found to emerge from decaying cowpats, the stacked dung of cattle, llamas and alpacas and all types of mixed dung. The latter also served as substrates for C. dewulfi and C. scoticus. In addition, C. scoticus was collected from cowpats and the dung of sheep and horses, whereas C. chiopterus and C. dewulfi were found in rotten cattle dung as well as in dung of mixed composition. Culicoides chiopterus developed quite frequently in cowpats and stacked cattle dung, horse dung and, to a lesser extent, in mixed sheep/horse dung, but also in duck feces. In summary, C. scoticus seems to be the Avaritia species that is least adapted to a specific kind of dung. Culicoides dewulfi was found at two of the studied dung sites only. It was the only species developing in chicken dung. Lower numbers of species, but higher numbers of specimens, of the subgenus Avaritia were recovered from llama and alpaca dung as compared to cattle or mixed dung. When individual sites are considered, the highest numbers of Culicoides species and specimens emerged from dung sites (sites 24, 29, 33, 36) and a renaturation area/floodplain (site 6) and the lowest number from meadows (site 5). In the subgenus Avaritia, high numbers of specimens were obtained from dung habitats (sites 32, 33, 36), floodplains (site 1) and compost (sites 24,26), and low numbers were obtained from swampy forests (sites 16,17) and boglands (site 23). The highest numbers of subgenus Culicoides specimens were collected in renaturation areas (site 6), mixed grassland/wetland A B Figure 2. Two-dimensional NMDS ordination plot of habitat preference by culicoid subgenus/group (A) and species (B-subgenus Avaritia excluded for species). Distance of habitats (black points) to biting midge subgenus and species (green triangles) are a measure of habitat preference. Abbreviations of landscape/habitat types: grass-/wetland without-mixed grassland/wetland without bushes and trees; grass-/wetland with-mixed grassland/wetland with bushes and trees.
For dung substrates, numbers of emerged specimens of the subgenus Culicoides were also much lower than those of the subgenus Avaritia, and no individuals of the subgenus Culicoides at all emerged from llama, alpaca, chicken and pig dung. Of C. pulicaris and C. punctatus, only single individuals could be found in rabbit, deer, sheep and duck dung and mixed goat/horse dung, and only a few more were found in mixed sheep/horse dung.
An NMDS test according to dung type was not possible due to insufficient data (stress value too small).
Muddy substrates used in renaturation areas, swampy forest areas, floodplains or meadows along streams were found to have a neutral pH value ( Figure 3A). By contrast, substrates from grassland/wetland and compost habitats (rotten vegetables) were slightly alkaline, while samples from dung were strongly alkaline. Overall, the higher occurrence probability of Avaritia and Culicoides species was associated with neutral or low pH values ( Figure 3A). The unexpected collection of culicoids from chicken, duck and pig dung indicates their potential to adapt to extreme breeding habitats with higher pH values ( Figure 3B). A total of 27.5% of the Culicoides samples collected by the emergence traps were males, 36.0% within the subgenus Avaritia, 16.6% within the subgenus Culicoides and 1.0% within other subgenera of the genus Culicoides. The highest percentage of C. obsoletus males was caught in September 2008 from a cattle dung heap, the highest percentage of C. chiopterus males in May 2008 from cowpats, the highest percentage of C. scoticus males in June and August 2009 from mixed cattle/horse dung and the highest percentage of C. dewulfi in May and September 2009 from cattle dung. In the Pulicaris Complex, the highest numbers of C. pulicaris and C. punctatus males were caught in floodplain areas and mixed grassland/wetland covered with bushes and trees in May and June 2008 and 2009, respectively.

Discussion
Despite recent extensive research on biting midge vectors, data on habitat binding and substrate preference by the various Culicoides species are still widely missing. One reason may be that data are almost impossible to generate by standardized methods and on a quantitative basis, since most biting midge species cannot be reared in the laboratory. Field data on breeding site productivity can rarely be quantitative and compared between locations and regions, even if the breeding sites are thoroughly characterized. Most likely, several additional unknown factors that determine productivity, not least the general composition of an area's biting midge fauna, are not taken into account when data are analyzed. Even with UV-light traps, commonly used for collecting adult biting midges, it is not possible to assess an area's general species composition satisfactorily, since not all species are attracted to these traps [47].
Moreover, UV-light traps commonly used for collecting adult culicoids [48][49][50][51][52] provide no information on Culicoides breeding sites or substrates. The above-mentioned limitations were partially overcome in the present study with the use of emergence traps, enabling the detection of Culicoides species freshly emerging from their developmental substrates. The advantages over manually collecting larvae from their substrates are manifold. Firstly, emerged adults can be morphologically identified much more easily than larvae or pupae, which are mostly difficult to identify or even unidentifiable at species level. Secondly, the emergence method is surface-related and can at least be standardized in this regard as the caught insects emerge from a well-defined area of the ground surface. Thirdly, it is not truly known how deep larvae wander into the ground in different habitats (e.g., under cowpats). Collecting potential breeding substrates from the ground surface might therefore miss developmental stages from below. Finally, the emergence trap technique is not invasive, so the soil biocenosis is not disturbed.
The study also showed that the CORINE land cover system, which is based on squares of 25 ha, is not appropriate for assigning potential breeding sites to landscape structures. Areas that are suitable or unsuitable as Culicoides breeding habitats differ on a much smaller scale. For the purpose of data analysis, the landscape structures in which traps were placed were therefore manually adjusted to real land use.
So far, 333 species of Ceratopogonidae have been described for Germany, with 57 of them belonging to the genus Culicoides [53,54]. In the present investigation, 21 Culicoides species were found in different German federal states by emergence traps, but the species diversity between sampling habitats and sites differed strongly. Some of the collected species seem to be ubiquitous, others to have a very limited and local distribution or to occur sympatrically with other species.
Species of the Obsoletus and Pulicaris Complexes occur almost worldwide [36] and are found all over Europe, whereby the taxonomic assignment of species to complexes (or "groups") is unclear and inconsistently practiced, resulting in tremendous difficulties in biological analyses [55,56]. Culicoides dewulfi, for example, is traditionally considered a member of the Obsoletus Complex due to its morphological similarity to C. obsoletus and C. scoticus, but its taxonomic status is still controversially discussed [57,58].
In the subgenus Avaritia, most of the female specimens collected in the present study were not identified to species level. The correlation between species and substrate was done on the morphological identification of males and the genetic identification of selected females. According to this, C. obsoletus, followed by C. chiopterus and C. scoticus, seems to be the species of the subgenera Avaritia, which accepts the most diverse breeding substrates. In the present investigation, these three species were caught at almost all sampling sites, especially often in floodplains, wetlands and the silting zones of lakes, and they were generally extremely abundant in all types of dung accumulations. González et al. [18] found species of the Obsoletus Complex in different types of manure and rotten organic material, while both Ninio et al. [20] and Thompson et al. [59] collected Obsoletus Complex specimens predominantly from cattle dung. Although Zimmer et al. [60] also detected C. obsoletus and C. scoticus in silage, these findings suggest that the dung of vertebrates is an important breeding habitat for the species of the subgenus Avaritia.
Regarding the variety of breeding habitats used, a similar situation is true for the subgenus Culicoides, which seems to occur ubiquitously in Germany [61][62][63]. Although the number of collected specimens of the subgenus Culicoides (mainly C. pulicaris and C. punctatus) was much lower than the number of specimens of the subgenus Avaritia in the present study, these species were found in almost all studied habitats, e.g., floodplains, grassland/wetland, banks of lakes, boglands and dung substrates (except manure). Thompson et al. [60] also noted species of the Pulicaris Complex in various types of dung, while González et al. [18] found C. punctatus to commonly develop in the mud of pond margins. In the present study, C. deltus developed in the same habitats as C. pulicaris and C. punctatus but, as opposed to those, had extremely high densities in renaturation areas. Culicoides newsteadi was found in meadows only.
Apart from the subgenus Avaritia, the most abundant and widespread group in the present study, all Culicoides species, including those of the Pulicaris Complex, were collected in small numbers. This may suggest that Culicoides species not belonging to the subgenus Avaritia are much more specialized and stenoecious regarding eco-climatic conditions. Although the data do not allow a founded phenological analysis, the collection data seem to suggest that the general emergence of culicoid species seems to peak twice annually, in May-June and in August-September. Differences in the abundance of the species of the subgenus Avaritia and of other subgenera occurred throughout the season, which is in agreement with the findings of Saerle et al. [64,65]. The species composition varied considerably between the sampled sites, with a clear tendency of numbers of species and specimens increasing with humidity and portion of decaying organic matter. It remains unclear, however, whether the low emergence rate in some habitats was a result of a low oviposition rate and, as a consequence, a low larval occurrence or of a combination of other biotic and/or abiotic factors.
The species C. achrayi, C. circumscriptus, C. festivipennis and C. nubeculosus are known to breed in organic-rich matter [66]. Hence, it is not surprising that these species were demonstrated in or alongside floodplains, river banks, reed belts, boglands and swampy areas. A similar investigation from Turkey [67] showed C. circumscriptus and C. festivipennis to be common in comparable habitats, while Uslu and Dik [68] found C. festivipennis, C. punctatus and C. brunnicans particularly often in mud at pond margins.
Observations that neutral or low pH values of substrates were associated with a higher occurrence probability of species of the subgenera Avaritia and Culicoides in contrast to alkaline substrates were already made by Steinke et al. [21] for C. obsoletus, C. chiopterus and C. dewulfi reared from cowpats. In this study, high pH values were also registered for the breeding substrates of C. nubeculosus and C. festivipennis. Both species had previously been found in slushy dung next to natural water reservoirs characterized by alkaline pH values [68,69]. Blackwell et al. [70] did not find any relationship between larval density and soil pH value for C. impunctatus in woodland and coastal habitats.
The present study tried to correlate landscape structures/habitats and breeding sites to subgenus and species, although the latter was not possible within the subgenus Avaritia due to the uncertain species assignment of the majority of females collected. The analysis shows a wide variability of breeding options accepted by many species, with a certain species-specific tendency of preferences. However, the data are still preliminary and should be supplemented by further studies.

Conclusions
Numerous studies have demonstrated that it is extremely difficult to define the requirements of biting midge species by their developmental substrates since their breeding habitats are widely unknown. Thus, even the identification of single emerging adults can be helpful to specify the adaptation of species.
Breeding habitats of putative Culicoides vector species, in which specimens develop in significant numbers, were identified for the subgenera Avaritia, mainly the Obsoletus Complex, in the form of dung substrates. Preferences of species in dung selection could be detected depending on the origin and the quality of the dung. Thus, in order to control the main vectors of BTV and SBV in Northern Europe which are considered to belong to the Obsoletus Complex [71,72], it may be an important step to properly manage dung deposition, storage and processing on farms. The breeding habitat preferences of other species including secondary vectors might affect landscape design and use.
Although this study has expanded the knowledge of habitats and dung substrates as Culicoides breeding sites, the selection and ecology of these are far from being defined. Further studies providing more detailed knowledge about the breeding sites and microhabitats of Culicoides biting midges in different landscape structures are urgently needed for the risk assessment of culicoid-borne diseases and targeted vector control.