@phdthesis{Scherer2019,
  author    = {Scherer, Philipp C{\´e}dric},
  title     = {Infection on the move},
  school      = {Universit{\"a}t Potsdam},
  pages     = {x, 107, XXXVIII},
  year      = {2019},
  abstract  = {Movement plays a major role in shaping population densities and contact rates among individuals, two factors that are particularly relevant for disease outbreaks. Although any differences in movement behaviour due to individual characteristics of the host and heterogeneity in landscape structure are likely to have considerable consequences for disease dynamics, these mechanisms are neglected in most epidemiological studies. Therefore, developing a general understanding how the interaction of movement behaviour and spatial heterogeneity shapes host densities, contact rates and ultimately pathogen spread is a key question in ecological and epidemiological research. In my thesis, I address this gap using both theoretical and empirical modelling approaches. In the theoretical part of my thesis, I investigated bottom-up effects of individual movement behaviour and landscape structure on host density, contact rates, and ultimately disease dynamics. I extended an established agent-based model that simulates ecological and epidemiological key processes to incorporate explicit movement of host individuals and landscape complexity. Neutral landscape models are a powerful basis for spatially-explicit modelling studies to imitate the complex characteristics of natural landscapes. In chapter 2, the first study of my thesis, I introduce two complementary R packages, NLMR and landscapetools, that I have co-developed to simplify the workflow of simulation and customization of such landscapes. To demonstrate the use of the packages I present a case study using the spatially explicit eco-epidemiological model and show that landscape complexity per se increases the probability of disease persistence. By using simple rules to simulate explicit host movement, I highlight in chapter 3 how disease dynamics are affected by population-level properties emerging from different movement rules leading to differences in the realized movement distance, spatiotemporal host density, and heterogeneity in transmission rates. As a consequence, mechanistic movement decisions based on the underlying landscape or conspecific competition led to considerably higher probabilities than phenomenological random walk approaches due directed movement leading to spatiotemporal differences in host densities. The results of these two chapters highlight the need to explicitly consider spatial heterogeneity and host movement behaviour when theoretical approaches are used to assess control measures to prevent outbreaks or eradicate diseases. In the empirical part of my thesis (chapter 4), I focus on the spatiotemporal dynamics of Classical Swine Fever in a wild boar population by analysing epidemiological data that was collected during an outbreak in Northern Germany persisting for eight years. I show that infection risk exhibits different seasonal patterns on the individual and the regional level. These patterns on the one hand show a higher infection risk in autumn and winter that may arise due to onset of mating behaviour and hunting intensity, which result in increased movement ranges. On the other hand, the increased infection risk of piglets, especially during the birth season, indicates the importance of new susceptible host individuals for local pathogen spread. The findings of this chapter underline the importance of different spatial and temporal scales to understand different components of pathogen spread that can have important implications for disease management. Taken together, the complementary use of theoretical and empirical modelling in my thesis highlights that our inferences about disease dynamics depend heavily on the spatial and temporal resolution used and how the inclusion of explicit mechanisms underlying hosts movement are modelled. My findings are an important step towards the incorporation of spatial heterogeneity and a mechanism-based perspective in eco-epidemiological approaches. This will ultimately lead to an enhanced understanding of the feedbacks of contact rates on pathogen spread and disease persistence that are of paramount importance to improve predictive models at the interface of ecology and epidemiology.},
  language  = {en}
}
@misc{HeinkenRaudnitschka2002,
  author    = {Heinken, Thilo and Raudnitschka, Dorit},
  title     = {Do wild ungulates contribute to the dispersal of vascular plants in central European forests by epizoochory?},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus-5850},
  year      = {2002},
  abstract  = {The external dispersal ("epizoochory") of vascular plant diaspores (seeds and fruits) by roe deer and wild boar, i.e. the most common wild large mammals with a large home range in central Europe, was investigated in a 6.5-km² forest area in NE Germany dominated by mesic deciduous forests. The study involved brushing out the diaspores from the coats and hooves of 25 shot roe deer and nine wild boar. The results were compared with the forest vegetation of the study area. Whilst wild boar transported large amounts of various diaspores in the coat, the significance of roe deer for epizoochory was low due to their sleek fur and different behaviour compared to wild boar. Altogether, 55 vascular plant species were transported externally. Since only a limited number of seeds came from woodland habitats, the open landscape was at least as important as a source of attached seeds as the forest vegetation. Thus, most plant species occurring in the studied forest area, especially characteristic woodland herbs, showed no adaptations to epizoochorous dispersal, although being very abundant in the herb layer. We conclude that hoofed game play a particular role concerning the dispersal of ruderal and grassland species in the agricultural landscape of central Europe. However, the actual spread of some herb species in forests of northern Germany, e.g. Agrostis capillaris, Brachypodium sylvaticum, Deschampsia flexuosa, Galium aparine and Urtica dioica, may be mainly facilitated by wild ungulates. Though dispersal by large mammals is an important mechanism for long-distance dispersal of plants in general, our results suggest that most of the characteristic herb species of mesic deciduous forests have only low epizoochorous dispersal potentials. The implications for nature conservation and silviculture are discussed.},
  language  = {en}
}