Population dynamics of rodents and their parasite communities in a naturally fragmented landscape

An island system with corresponding mainland sites, was used to study woodland rodent dynamics and their parasite communities within a naturally fragmented landscape. The study site, hosts and parasite species investigated within this thesis allowed the investigation of how natural fragmentation affects demographic and population dynamics of rodents (chapter 3). Reduced habitat connectivity is known to affect nearly every process in biology. Low degrees of fragmentation and high connectivity between habitats have been shown to provide the most stable conditions for populations to persist, as movement of organisms is less restricted. It is shown that in contrast to previous studies on fragmented populations, the fragmented landscape of the islands had little effect on the demographic characteristics of rodent populations in comparison to those on the mainland. There were few difference found in the demographics of wood mice and bank voles when compared to mainland sites. The results from this study then allowed the broader question of how parasites dynamics are affected by the spatial structure of a host population to be addressed. Theory predicts that parasites are unable to persist in small, isolated host populations, due to small host population size as well as potential genetic factors increasing the risk of extinction. However parasites may become more prevalent in isolated populations as hosts may have a reduced ability to deal with infection. It is shown (chapter 4) that within this study system that despite some island populations being extremely small, there is no overall reduction in parasite species found within fragmented habitats. Furthermore, extinction of the parasites investigated within wood mice and bank voles is unlikely due to the direct life cycle of these parasites. Variation was seen in the prevalence of infection, however the majority of the parasite species on islands did not show a reduced prevalence of infection compared to mainland sites. Finally parasite co-infection and co-aggregation and their dependency on host characteristics in woodland rodents (chapter 5) were investigated. Parasite species infecting hosts are normally studied individually, however this is not what is seen within natural populations. Co-infection is an important concept within natural systems as there is a vast diversity of parasite species that create ample opportunity for concurrent infections. Therefore, it is proposed that studies should be focused on parasite interactions, as within host interactions can in turn affect the abundance and distribution at the level of the host population. This study focused on seven parasite taxa, and it was found that the maximum number of parasite species any individual was found to be infected with was five, with the mean number for both host species at around two. Parasites associations were also more common than expected within the same functional groups with co-occurrence being more common between parasite species associated with ectoparasites. Within this study, host aggregation was positively correlated with differing parasite taxa. Furthermore, looking at patterns of co-aggregation could aid in our understanding of parasite interactions within hosts. The nature of these interactions will determine whether aggregation is positively or negatively correlated across different parasite taxa. A small number of hosts maybe responsible for transmitting the majority of infections (20/80 rule). Identifying these individuals would be informative in helping to control disease spread. Host characteristics have been found to be informative in terms of single parasite species infections. Within this study it was found that juvenile bank voles were more likely to be co-infected than those within other age classes. No host characteristic explained patterns of co-infection in wood mice. In conclusion I found that natural fragmentation does not have an overall negative effect on rodent host dynamics nor does it reduce the number or prevalence of infection of parasite species able to infect hosts. This thesis has highlighted the importance for using natural wildlife systems in empirical studies, and the need to further address multiple parasite interactions within a host community