Introduction
Avian malaria (Plasmodium) and malaria-like (Haemoproteus) parasites (Protozoa, order Haemosporidia) are abundant on all continents, except Antarctica (Valkiūnas, Reference Valkiūnas2005) and the infections caused by these parasites can have serious physiological impacts on birds (Atkinson and van Riper, Reference Atkinson, van Riper, Loye and Zuk1991; Merino et al., Reference Merino, Moreno, Sanz and Arriero2000; Knowles et al., Reference Knowles, Palinauskas and Sheldon2010; Asghar et al., Reference Asghar, Hasselquist, Hansson, Zehtindjiev, Westerdahl and Bensch2015). The effects of these parasites can be particularly strong for individuals of immunologically naïve host species and the arising negative fitness consequences may affect entire populations (van Riper et al., Reference van Riper, van Riper, Goff and Laird1986; Townsend et al., Reference Townsend, Wheeler, Freund, Sehgal and Boyce2018).
The advent of molecular techniques in blood parasite detection revealed a remarkable species richness, diversity and distribution range of these parasites. To date we know of 2303 Plasmodium and Haemoproteus lineages detected from 1479 bird species (Malavi database: http://mbio-serv2.mbioekol.lu.se/Malavi); however, the number of different parasite lineages is thought to exceed 10 000 (Bensch et al., Reference Bensch, Pérez-Tris, Waldentsröm and Hellgren2004). Consequently, studies investigating the costs of avian malaria and malaria-like infections need to identify lineages involved in the given host–parasite relationship.
A large variation in avian Haemosporidian diversity and prevalence was found between, but also within a host species, e.g. when inhabiting different geographical regions (e.g. Hellgren et al., Reference Hellgren, Pérez-Tris and Bensch2009; Marzal et al., Reference Marzal, Ricklefs, Valkiūnas, Albayrak, Arriero, Bonneaud, Czirják, Ewen, Hellgren, Horakova, Iezhova, Jensen, Križanauskienė, Lima, de Lope, Magnussen, Martin, Møller, Palinauskas, Pap, Pérez-Tris, Sehgal, Soler, Szöllősi, Westerdahl, Zetindjiev and Bensch2011; Szöllősi et al., Reference Szöllősi, Cichoń, Eens, Hasselquist, Kempenaers, Merino, Nilsson, Rosivall, Rytkönen, Török, Wood and Garamszegi2011). Environmental conditions can influence the prevalence and virulence of different parasite species, both through the abundance and diversity of vector populations and also through the composition of bird host species communities (Scheuerlein and Ricklefs, Reference Scheuerlein and Ricklefs2004; Valkiūnas, Reference Valkiūnas2005; Hellgren et al., Reference Hellgren, Waldenström, Pérez-Tris, Szöllősi, Hasselquist, Krizanauskiene, Ottoson and Bensch2007; Soltész, Reference Soltész2017). In addition, each parasite species has its own transmission period, possibly causing variances in parasite prevalence patterns in different areas and seasons (Valkiūnas, Reference Valkiūnas2005).
The prevalence and virulence of a parasite species may also vary due to mutual adaptations between vectors, birds and the parasites. For example, vector species may differ in their host breadth so that some of the vectors have wide host preference and suck blood from a range of bird species with a diverse array of parasite species while others are host specialists (Valkiūnas, Reference Valkiūnas2005; MalAvi). Similarly, the parasites may also vary in their ability to complete their development in a range of distantly related vector and bird host species (generalist parasites) or only in a smaller group of closely related species (specialist parasites). Further, the strength of the immune response of the hosts affects the prevalence and intensity of parasite infections (reviewed in Westerdahl, Reference Westerdahl2007; Marzal et al., Reference Marzal, Møller, Espinoza, Morales, Lujan-Vega, Cardenas-Callirgos, Mendo, Alvarez-Barrientos, Gonzalez-Blazques, Garcia-Longoria, de Lope, Mendoza, Iannacone and Magallanes2018; Antonides et al., Reference Antonides, Mathur, Sundaram, Ricklefs and DeWoody2019).
Avian malaria and malaria-like parasites show a complex life cycle. They reproduce asexually in birds (intermediate hosts) and sexually in insect vectors (definitive hosts; vectors of Plasmodium are species from the family Culicidae; vectors of Haemoproteus are species from the family Ceratopogonidae and Hipposcidae). Larvae of the vectors of these parasites develop in a wide variety of aquatic habitats (Soltész, Reference Soltész2017) and freshwater wetlands with reed beds particularly represent an ideal environment because of the high humidity and low salinity of the habitat (Cox, Reference Cox1993). These conditions may result in high vector abundance and diversity, and in turn, a high prevalence of blood parasites in birds nesting in these habitats.
In this study, we investigated a specialized reed dwelling species, the bearded reedling Panurus biarmicus. They spend their entire life in freshwater wetlands, and considering the optimal developmental conditions for the vector species, we predicted a very high risk of blood parasite infestation in bearded reedlings. Consequently, we expected a high prevalence as well as high parasite diversity.
Previous studies focused mainly on the description of the parasite fauna of various bird communities breeding in freshwater wetlands (Ventim et al., Reference Ventim, Tenreiro, Grade, Encarnacao, Araújo, Mendes, Pérez-Tris and Ramos2012a, Reference Ventim, Morais, Pardal, Mendes, Ramos and Pérez-Tris2012b; Neto et al., Reference Neto, Pérez-Rodríguez, Haase, Flade and Bensch2015; Nilsson et al., Reference Nilsson, Taubert, Hellgren, Huang, Palinauskas, Markovets, Valkiūnas and Bensch2016; Dimitrov et al., Reference Dimitrov, Ilieva, Ivanova, Brlík and Zehtindjiev2018). However, less is known about the parasites that are transmitted locally (but see Ventim et al., Reference Ventim, Tenreiro, Grade, Encarnacao, Araújo, Mendes, Pérez-Tris and Ramos2012a, Reference Ventim, Mendes, Ramos, Cardoso and Pérez-Tris2012c; Nilsson et al., Reference Nilsson, Taubert, Hellgren, Huang, Palinauskas, Markovets, Valkiūnas and Bensch2016; Dimitrov et al., Reference Dimitrov, Ilieva, Ivanova, Brlík and Zehtindjiev2018) or about the consistency of parasite communities across populations of a host species.
For this reason, in our descriptive study, we investigated two geographically distinct bearded reedling populations and compared our findings with parasite data reported previously in other wetland passerine populations.
Materials and methods
Study species and field methods
The fieldwork was carried out at Hortobágy-Halastó (47°37′N, 21°04′E, Hungary) and at Lake Neusiedl (47°56′N 16°46′E, Austria) (Fig. 1). The reed bed area in Hortobágy-Halastó is 4.52 km2 and in Lake Neusiedl is 180 km2 in size. The bearded reedling is a small (~13 g), resident, wetland specialist passerine common throughout Eurasia. It is dependent on extended reed beds throughout its life, and uses this habitat for breeding, foraging and wintering (Cramp and Perrins, Reference Cramp and Perrins1993). We caught a random sample of 47 male and female individuals with mist nets between mid-April and mid-July 2016 in three sampling points in Hungary and 51 male and female individuals between mid-August and mid-October 2016 in two sampling points in Austria (Fig. 1). Individuals sampled in the Hungarian population consisted of both adults (n = 35, sampled from mid-April onwards) and juveniles (n = 12, sampled from mid-June onwards), whereas most individuals sampled in Austria had already completed their post-juvenile moult and were in adult plumage by the time of sampling.
Sampling points of bearded reedlings (Panurus biarmicus). We collected blood samples from two distinct populations, one from Hungary, Hortobágy-Halastó (n = 47 individuals from three sampling points, HU1, HU2 and HU3) and one from Austria, Lake Neusiedl (n = 51 individuals from two sampling points, AT1 and AT2).
Sample collection and parasite detection
Blood samples were taken from the brachial vein and stored on ice in the field and separated to plasma and cell fractions (used in the present analysis) by centrifuging the samples on 2200 g for 5 min. Thereafter the samples were kept at −20 °C until further analyses in the laboratory. DNA was extracted by an ammonium-acetate method (Nicholls et al., Reference Nicholls, Double, Rowell and Magrath2000) then the concentration of genomic DNA was adjusted to ca. 35 ng µL−1 by quantifying genomic DNA on a BioTek Epoch Microplate Spectrophotometer. A highly sensitive polymerase chain reaction (PCR) based method was used for the detection of parasite prevalence (Waldenström et al., Reference Waldenström, Bensch, Hasselquist and Östman2004). In all PCRs, both negative (ddH2O) and positive controls (DNA samples from individuals of other bird species that were previously confirmed to be infected) were used to control for possible contaminations and failures during PCRs, respectively. None of the negative controls ever showed any amplification of parasite DNA, and parasite DNA was always amplified correctly from positive controls. To ascertain that none of the samples went through degradation between sample collection and parasite investigation and that the samples were in a good condition for PCR analyses, all samples were checked for DNA quality by amplifying the CHD (chromo-helicase-DNA-binding) genes of bird host DNA (using the primer pairs described by Griffiths et al., Reference Griffiths, Double, Orr and Dawson1998). To reduce the risk of losing infections because of sampling error, negative samples were screened twice for blood parasites.
All samples with positive amplification for parasite DNA were sequenced directly using the BigDye® Terminator v3.1 cycle sequencing kit. Products from the sequencing reactions were purified and sent to a capillary electrophoresis platform (Biological Research Centre, Hungary). Sequences were edited and aligned using the program BioEdit (Hall, Reference Hall1999) and identified to genus (Haemoproteus or Plasmodium) and lineage level by comparing sequence data with those of the previously identified parasites reported in MalAvi database.
Data analyses
Mean prevalence (No of infected individuals/No of tested individuals × 100) of Haemoproteus and Plasmodium parasites in each host species; Species evenness (E) and Simpson's index (D) for measure parasite diversity in each host population were calculated. Species evenness (E) was calculated from the formula E = D/D max (where D max means the number of parasite species/population). Simpson's index (D) was calculated using the formula as follows (where p i is the relative abundance of the lineages in a population):
$$D = \displaystyle{1 \over {\sum\limits_{i = 1}^S {\,p_i^2}}}. $$The few sampled populations combined with the number of effects to control for (i.e. species, populations, migratory status, locality, phylogeny) prevented us to perform statistical analyses on the indices describing prevalence patterns in the different species.
Results
Detected lineages and parasite prevalence
None of the 47 Hungarian bearded reedling samples and only two out of the 51 Austrian samples showed a positive amplification for either Plasmodium or Haemoproteus. This means 0 and 3.9% prevalence in the Hungarian and the Austrian populations, respectively. Both birds that showed positive amplification for malaria in the Austrian population were infected with the same Plasmodium lineage, pSGS1 (see also Table 1). Mean prevalence; Species evenness (E) and Simpson's index (D) are listed in Table 1.
Overall Haemoproteus and Plasmodium prevalence (percentage of infected individuals) and parasite lineages detected in different populations of bearded reedlings and coexisting wetland breeding passerines
Only species with at least two populations with a minimum of 40 sampled individuals (i.e. the lowest number of sampled individuals in the bearded reedling populations) and with exact sampling locations are shown. The symbols after certain lineages (e.g. ●, ■, ♥) indicate that the lineage was detected in more than one population of the given species. Lineages shown in bold mean that they would appear also in other population(s) of the given species, but because of the low sample sizes those populations are not indicated in the table. * Indicates that in the Kvismaren population of the great reed warbler some of the individuals were sampled more than once (No of tested samples = 782, No of sampled individuals = 470). This may cause a bias in the estimates of overall prevalence and in the lineage prevalence; however, survival was unrelated to infection status of the individuals in the given species.
Mean prevalence of Haemoproteus and Plasmodium parasites for each host species, Simpson's index (D) and Species evenness (E) for the parasites were also calculated in each host population.
Discussion
We found extremely low Haemosporidian prevalence and diversity in the two distinct bearded reedling populations in Hungary and Austria. Prevalence in the Hungarian population was 0%, while in the Austrian population only two out of the 51 sampled individuals (3.9%) were infected with avian malaria and with only one Plasmodium lineage (pSGS1). This consistent and remarkably low infection level is in contrast to our predictions namely that due to the environment, prevalence, as well as diversity, should be high. This result hence suggests that bearded reedlings are either very sensitive to malaria causing the selective disappearance of infected individuals or that the species has special adaptations for malaria resistance.
We can rule out the possibility that the low prevalence found is due to the weaknesses of the techniques used in parasite detection. We used a highly sensitive PCR-based method that can detect infections in as low intensities as 0.001% and sometimes in intensities as 0.0001% (Waldenström et al., Reference Waldenström, Bensch, Hasselquist and Östman2004) which is much below the detected parasitaemia in wild-caught birds during the breeding season (0.01–1%, Valkiūnas, Reference Valkiūnas2005). In addition, this method has already proved to successfully detect various Haemosporidian lineages from different bird species also in our laboratory (e.g. from collared flycatchers, see Szöllősi et al., Reference Szöllősi, Garamszegi, Hegyi, Laczi, Rosivall and Török2016, Jones et al., Reference Jones, Kulma, Bensch, Cichoń, Kerimov, Krist, Laaksonen, Moreno, Munclinger, Slater, Szöllősi, Visser and Qvarnström2018; from house sparrows, see Marzal et al., Reference Marzal, Ricklefs, Valkiūnas, Albayrak, Arriero, Bonneaud, Czirják, Ewen, Hellgren, Horakova, Iezhova, Jensen, Križanauskienė, Lima, de Lope, Magnussen, Martin, Møller, Palinauskas, Pap, Pérez-Tris, Sehgal, Soler, Szöllősi, Westerdahl, Zetindjiev and Bensch2011). Furthermore, all negative samples were screened twice and our positive controls were always amplified correctly from the samples. We also used a reasonable sample size for both populations (n = 47 and 51 in the Hungarian and the Austrian populations, respectively). Though the sampling period was different for the Hungarian (samples were collected between mid-April and mid-July 2016) and the Austrian populations (samples were collected between mid-August and mid-October 2016), both populations were screened during the most active period of the vectors. Though the Hungarian population consisted of both juvenile and adult individuals, also juveniles in the Hungarian population had time to develop infections as they were sampled from June onwards, while the first fledglings appear in early April. This makes it unlikely that differences in the sampling periods would have been responsible for the pattern reported. We therefore suggest that the genuinely low prevalence in our populations has an ecological explanation.
Parasite prevalence varies considerably among host species and sometimes also between different populations of a species. One straightforward explanation for these differences in Haemosporidian prevalence is that the abundance and the diversity of the suitable vectors differ between locations due to differences in local climatic and environmental conditions necessary for vector and parasite development (Burkett-Cadena et al., Reference Burkett-Cadena, McClure, Estep and Eubanks2013; Golding et al., Reference Golding, Nunn and Purse2015).
From the year of sampling, we have information about the mosquito fauna (family Culicidae) of the Austrian study site (Führer et al., unpublished data), while mosquitoes at the Hungarian site were sampled in previous years (in 2011 and in 2012, Soltész, Reference Soltész2017). These studies show that both sites harbour a vast array of blood sucking vectors. The mosquito species found at Hortobágy-Halastó included Culex modestus, C. pipiens and Ochlerotatus caspius from which altogether more than 40 Haemoproteus and Plasmodium lineages were detected, out of which 21 were also proved to be transmitted in Europe. At Lake Neusiedl, Führer et al. reported the presence of C. modestus, C. pipiens and Uranotaenia unguiculata (unpublished data). From these species 42 Haemoproteus and Plasmodium lineages were detected and out of which 22 were proved to be transmitted in Europe (MalAvi). These studies and our personal experience in the field sites suggest that a highly diverse and abundant vector fauna is available for parasite transmission. Based on the potentially available vector and parasite species, one would expect a high rather than low blood parasite prevalence and diversity in the sampled bearded reedling populations.
However, none of the sampled birds was infected with either Haemoproteus or Plasmodium at Hortobágy-Halastó, and the prevalence and diversity were also very low in the Austrian population (3.9% prevalence, D = 1). These results are consistent with the low parasite prevalences and diversities reported from other populations of bearded reedlings (Table 1). Among the three populations tested previously across Europe, at Lake Durankulak, Bulgaria, two out of the 35 adult and none of the 18 juvenile individuals sampled during spring and summer (in year 2006–2008, 2013–2016) were detected to be infected with Haemosporidian parasites, making a 3.8% overall prevalence in this population. However, for some reason, the authors were not able to sequence their samples; therefore, no information is available about the lineages infecting this population (Dimitrov et al., Reference Dimitrov, Ilieva, Ivanova, Brlík and Zehtindjiev2018). Similarly, in a Swedish population at Lake Kvismaren, where Nilsson et al. (Reference Nilsson, Taubert, Hellgren, Huang, Palinauskas, Markovets, Valkiūnas and Bensch2016) screened juvenile bearded reedlings during summer (year 2004), only four out of the 40 individuals were proved to be infected with Haemoproteus. Out of these four individuals, three were infected with hWW2 (7.5% prevalence) and one with hTURDUS2 (2.5% prevalence). In addition, in a Turkish population of the species sampled at Sultan Marshes, Yildirim et al. (unpublished) screened altogether 58 adult and juvenile individuals and found that five out them were infected with hGRW5 (8.6% prevalence) and one with hWW1 (1.7% prevalence) (see also MalAvi database). While zero prevalence in our Hungarian population is the lowest reported, consistent reports of low infection levels suggest that low Haemosporidian prevalence is a general phenomenon in this species.
Some previous studies also found low Haemosporidian prevalence in other bird species. However, these species were sampled in habitats where vectors were either absent or the conditions for parasite development were not favourable [i.e. bird species living on islands far from the continents (Campioni et al., Reference Campioni, Martínez-de la Puente, Figuerola, Granadeiro, Silva and Catry2018); on coastal areas (Martínez-de la Puente et al., Reference Martínez-de la Puente, Munoz, Capelli, Montarsi, Soriguer, Arnoldi, Rizzoli and Figuerola2015); or in high altitudes (Valkiūnas, Reference Valkiūnas2005)]. Low parasite prevalence and/or diversity of parasites were also found in introduced bird species, which had ‘lost’ their parasites during the colonization process (Lima et al., Reference Lima, Simpson, Fecchio and Kyaw2010; Ventim et al., Reference Ventim, Mendes, Ramos, Cardoso and Pérez-Tris2012c). Similar findings apply to host species characterized by small population sizes and fragmented distribution, preventing the maintenance of specialized parasites (Neto et al., Reference Neto, Pérez-Rodríguez, Haase, Flade and Bensch2015; Eastwood et al., Reference Eastwood, Peacock, Hall, Roast, Murphy, Gonçalves da Silva and Peters2019). Remarkably low Haemoproteus prevalence was reported from the chiffchaff species complex (Phylloscopus collybita), although the sample sizes in the sampled populations were moderate (5–30), therefore the prevalence of parasites could not be estimated precisely (Bensch et al., Reference Bensch, Jönsson and Copete2012).
If a bird species is utilized by host-specific parasites, the transmission may be hampered by very low population sizes or if the population goes through a bottleneck. Though we have no information about the exact population sizes of the bearded reedlings sampled across Europe, previous bird census data showed that the Hungarian study population consists of ca. 1200 individuals (Lendvai et al., unpublished data) and the Austrian population of approximately 5000–10 000 individuals (Hoi et al., unpublished data). These population sizes cannot be considered to be small. However, we also know from previous studies that the species might be sensitive to the low temperatures and high precipitation during winter, which strongly affect the population size in the following year (Lendvai et al., unpublished data; Wilson and Peach, Reference Wilson and Peach2006). This may result in strong fluctuations in the population sizes from year to year which may perhaps provide an unfavourable environment for host specialized parasites and thus results in a lack of such parasites in this species (Neto et al., Reference Neto, Pérez-Rodríguez, Haase, Flade and Bensch2015).
Indeed, neither of the bearded reedling populations studied to date has any common blood parasite. The Plasmodium lineage infecting the Austrian population (pSGS1) is highly host generalist, detected previously in 123 other bird species. Also the parasites infecting the previously studied Swedish (hWW2, hTURDUS2) and Turkish (hGRW5, hWW1) populations were detected from a range of 4–21 bird species (MalAvi). The least host generalist among these lineages is the lineage hGRW5 that was detected from three other bird species, however, these bird species are all Acrocephalus species (Acrocephalus arundinaceus, A. melanopogon, A. palustris) coexisting in the reed beds with the bearded reedling. The lack of common bearded reedling specialist parasites and the generally low overall prevalence of parasites in this species raise the possibility that the parasites infecting the bearded reedling populations may occasionally come from other coexisting bird species of the sampling areas and that these parasites cannot develop in this bird species. The idea that bearded reedlings may be unsuitable hosts for avian malaria and malaria-like parasites is supported by the fact that prevalence of blood parasites in the coexisting wetland specialist bird species was much higher (Table 1). Even more, each of these other bird species carried some parasite lineages that were present in more than one population of the given species (Table 1) suggesting that these lineages can indeed develop in these bird species.
One caveat though is that all these coexisting species are migratory, because sufficient information (i.e. a reasonable number of individuals/populations) on parasite prevalence was not available for resident wetland specialists (MalAvi). This means that higher infection patterns in coexisting migratory species could also be explained by the fact that these species are exposed to the vectors of Haemosporidian parasites not only during breeding but also during migration and at their wintering sites (e.g. Waldenström et al., Reference Waldenström, Bensch, Kiboi, Hasselquist and Ottosson2002; Lemke et al., Reference Lemke, Tarka, Klaassen, Åkesson, Bensch, Hasselquist and Hansson2013; Hegemann et al., Reference Hegemann, Abril, Muheim, Sjoberg, Alerstam, Nilsson and Hasselquist2018).
However, high overall Haemosporidian prevalence was also detected in different populations of resident bird species from habitats other than wetlands. In blue tits (Cyanistes caeruleus) (Szöllősi et al., Reference Szöllősi, Cichoń, Eens, Hasselquist, Kempenaers, Merino, Nilsson, Rosivall, Rytkönen, Török, Wood and Garamszegi2011) and great tits (Parus major) (MalAvi database) from natural forests and urban areas, the average prevalence was much higher [mean prevalence 66.3% (ranging from 30.5 to 100%) and 59.7% (ranging from 32.6–90.0%), respectively] than in bearded reedlings, even though the abundance and diversity of vectors in those habitats were probably lower than in wetlands. Blue tits and great tits were also infected with a range of parasite lineages commonly found in each population (1–5 and 2–3 common lineages/blue tit and great tit populations, respectively), suggesting that host–parasite coevolutionary relationships are probably more important in determining parasite prevalence and diversity than environmental factors. This reasoning is supported by a recent study, where certain parasite lineages were strongly associated with the different host species and prevalence of Haemosporidian parasites was more similar between different populations of a species than across species within a habitat (Pulgarín-R et al., Reference Pulgarín-R, Gómez, Robinson, Ricklefs and Cadena2018).
It is therefore possible that bearded reedlings are resistant to avian malaria and malaria-like infections and either do not become infected or quickly clear occasional infections. This might either be because bearded reedlings are immunologically adapted or resistant to these infections or apply some unknown behavioural strategies that make them less likely to be exploited by the vectors of Plasmodium and Haemoproteus parasites. It is also possible that infections by certain pathogens may exclude or reduce the possibility of infections by others (see e.g. Richie, Reference Richie1988; Nacher, Reference Nacher2004).
The low prevalence and parasite diversity in bearded reedling populations may also indicate that mortality among infected individuals is so high that this manifests in no/only a few infected individuals detected in the population. However, such a high mortality (especially before the first reproduction) should cause very strong selection towards Haemosporidian resistance. To answer these possibilities, further infection experiments and studies on resistance genes are needed. In any case, the remarkably low infection level of bearded reedlings makes them a useful candidate model species for either behavioural or immunological resistance or extreme sensitivity towards avian blood parasite infections.
Acknowledgements
We thank Orsolya Vincze for her help in the field and the support of the Hortobágy National Park.
Financial support
The study was supported by the National Research, Development and Innovation Office (OTKA K113108 and TÉT 15-1-2016-0044) to Á.Z.L. and (PD124043, FK127917) to E.S., the Bolyai János research fellowship (bo_50_17) to E.S., and the Hungarian Ministry of National Resources-National Talent Program (NTP-EFÖ-P-15-041) to Z.T. the European Union and the European Social Fund (EFOP-3.6.1-16-2016-00022) and by the Hungarian Eötvös Scholarship (MÁEÖ2016_15/76740 to O. Vincze). During the preparation of the manuscript Á.Z.L. was supported by the Romanian Ministry of Education (PN-III-P4-ID-PCE-2016-0572).
Conflict of interest
None.
Ethical standards
The present study complies with the current laws of Hungary and Austria. This study was done with permissions from the regional government agency (HBB/17/00870-3/2015) in Hungary. In Austria, all animal procedures were carried out with the approval of the institutional ethics committee and the national authorities according to §26 of the Law for Animal Experiments, Tierversuchsgesetz 2012 – TG 2012 (permit number: GZBMWF-68.205/0012-WF/V/3b/2017).
