Dokument

Summary:
The global biodiversity crisis is advancing steadily. Additionally, climate change increases the risk of emerging infectious diseases (EIDs). Due to their global migration, birds are of special interest when studying the spread of pathogens and parasites. Avian influenza, Trichomonas gallinae, and the Usutu virus are only some examples of EIDs that had significant negative effects on their host species populations. However, also more less prominent parasites and pathogens are expected to shift their distribution ranges in response to climate change, such as avian malaria and other haemosporidian parasites. While infection with haemosporidian parasites is usually not lethal in co-evolved host species, host-switching events have had severe negative effects on the new host species, even leading to local extinctions of some species. Given their importance in understanding emerging infectious diseases (EIDs) and documented population declines linked to environmental factors, birds play a crucial role in studying the impact of pathogens, parasites, and environmental changes on populations and communities. While population declines are often only recognized when the number of individuals decreases, physiological measures offer the potential to predict declines and identify their causes. For instance, assessing stress levels in birds, which is associated with fitness loss, can serve as an early warning indicator, particularly when combined with monitoring for parasites. One method for assessing the physiological stress response in birds is by using the ratio of heterophils to lymphocytes (H/L ratio) which has been found to increase in response to increased stress hormones. Both physiological stress and the effects of parasites have primarily been studied within single species. However, there is a lack of research on bird communities. In my doctoral thesis, I delved deeper into stress and endoparasites within a European forest bird community. My aim was to determine the suitability of analysing these factors as warning indicators across multiple species and to assess the influence of environmental factors on them. To achieve this, I studied the prevalence, diversity, and host-parasite interactions of endoparasites and their hosts within the community. Additionally, I contributed to the development of a novel deep learning approach for automatic avian blood cell identification and counting. Furthermore, I evaluated the natural variation of the H/L ratio across the bird community, considering factors such as phylogeny, sex, age, body condition, and incubation status. Lastly, I investigated the impact of temperature, precipitation, and shrub layer, as well as endoparasite infection, on the H/L ratio of the forest bird community. For data collection, I captured and sampled 483 bird individuals belonging to 29 species in a managed forest in Central Germany over four breeding seasons. I collected saliva and blood samples to examine infections with endoparasites, with a focus on T. gallinae and haemosporidian blood parasites. Additionally, the blood samples were used to determine the H/L ratio from a blood smear. In my research, I found that nearly half of the sampled birds (48.1%) were infected with blood parasites. However, these infections were not evenly distributed across species; some species, such as Parus major and Turdus merula, exhibited significantly higher susceptibility to infections compared to others. For instance, no infections were found in species such as Certhia familiaris and Dendrocopos major. To improve blood smear analysis and ensure consistent counting, I contributed to developing a deep learning method for identifying and counting bird blood cells. It involved two neural networks: one for selecting countable regions with a 97.3% F1 score (the harmonic mean of precision and recall), and another for identifying and counting cells within these regions, achieving a mean average precision of 90.7%. Based on the resulting count data I calculated the H/L ratio and detected a strong phylogenetic signal across the studied forest bird community. I further found general patterns of the influence of intrinsic factors on the H/L ratio within the community. Females had significantly higher H/L ratios than males and adult birds had higher H/L ratios than females and adults had higher H/L ratios than juveniles. We further found that birds involved in the incubation of eggs tended to have higher H/L ratios than non-incubating birds. Body condition did not significantly affect the H/L ratio across the bird community. These findings emphasise the importance of controlling for intrinsic factors when using the H/L ratio as an indicator. Finally, I also included abiotic factors (temperature and precipitation) and biotic factors (shrub layer) in my analyses and found that higher mean decade temperature had a significant negative impact on the H/L ratio. The results further indicate that infected birds respond differently to environmental factors than uninfected ones. The stress level in infected birds does not change with temperature, and while uninfected birds tended to be less stressed at capturing sites with denser shrub layer, infected birds show contrasting trends. Based on my analyses, I demonstrated that both stress and parasite infestation can be evaluated across species in a forest bird community and hence potentially serve as early warning indicators for bird communities especially since both factors have been associated with reduced reproductive success and reduced survival in previous studies. However, future studies in bird communities should also incorporate additional factors such as flight activity and reproductive success to confirm the validity of stress and parasite infestation as early warning indicators.

Das Dokument ist im Internet frei zugänglich - Hinweise zu den Nutzungsrechten