The ongoing loss of global biodiversity is unprecedented in both magnitude and pace, raising urgent questions as to how this loss will affect ecosystem functioning and human well-being. Control of infectious diseases has been proposed as an important ecosystem service that is likely to be affected by biodiversity loss. A negative relationship between biodiversity and disease risk could offer a win-win situation for nature conservation and human health. However, the generality of this relationship remains the subject of contentious debate. The aim of this thesis was to contribute to a better understanding of the interactions between ticks and their vertebrate hosts in a biodiversity hotspot, and how loss of biodiversity affects these interactions and ultimately, tick-borne disease risk. My study was unique in that I simultaneously considered and directly assessed broader communities of Neotropical wildlife, ticks, and tick-borne pathogens across an anthropogenic disturbance gradient. Determining whether and how biodiversity loss affects tick-borne disease risk in tropical forests requires a thorough understanding of tick-host associations, which are a function of tick-host specificity as well as host biological and ecological traits. In chapter 2, I therefore quantified the degree to which adult ticks are host-specific in my study region: Panama. Using quantitative network analyses and phylogenetic tools with null model comparisons, I found that the adult life stages of most tick species were specific to a limited number of host species that were phylogenetically closely related. In Chapter 4 I showed that species assemblages of adult ticks became increasingly diverse on larger-bodied host species, indicating that adult ticks in Panama tend to select for large reproduction hosts. In contrast to adult ticks, understanding the ecological interactions between immature ticks and their hosts in the tropics has long been hampered by a lack of morphological identification keys. Therefore, in Chapter 3, I describe the development of a DNA barcode reference library for the molecular identification of larvae and nymphs. This reference library was highly effective in species-level identification of immature ticks collected from birds (Chapter 3) and small mammals (Chapter 4 and 6). Several avian ecological traits were positively associated with tick parasitism, but the potential role of wild birds in tick-borne disease transmission seems to be limited in Panama. Immature ticks did not show any specificity to particular bird species or avian ecological traits (Chapter 3), and species assemblages of immatures ticks were equally diverse across a large number of host taxa (Chapter 4). This suggests that larvae and nymphs may feed more opportunistically than their adult counterparts. High host specificity in adult ticks implies high susceptibility to tick-host coextinction, even if immature ticks feed opportunistically. In chapter 5, I tested this hypothesis by surveying tick and vertebrate host communities across a forest fragmentation gradient. Forest fragments consisted of previously connected islands and peninsulas in the Panama Canal and ranged 1000-fold in size. Abundance and species richness of ticks was positively related to that of wildlife, which in turn was related to the size of the forest fragment. Specialist tick species were only present in fragments where their specific reproduction hosts were captured by camera traps. Further, less diverse tick communities were dominated by a generalist tick species. These results indicate that loss of wildlife had cascading effects on tick communities through local host-parasite coextinction. In Chapter 6, I studied how communities of wildlife, ticks, and tick-borne microbes changed along a more ‘typical’ disturbance gradient, in which forest fragments were embedded in an agricultural and sub-urban landscape, rather than surrounded by water. I found that wildlife community disassembly either diluted, amplified, or had no effect on infection prevalence in ticks, depending on the pathogen and degree of disturbance. However, hyperabundance of medium- to large-sized frugivores and herbivores (important reproduction hosts for adult ticks) in sites that lacked apex predators was related to exponential increases in tick density, negating any effect of reduced pathogen prevalence. Moreover, high tick species richness in these sites was related to high microbial and pathogen richness. High parasite diversity is thus a source of infectious diseases. When medium- to large-sized frugivores and herbivores also disappeared, densities of infected ticks declined, suggesting a non-linear relationship between biodiversity loss and tick-borne disease risk, in which initial loss of apex predators increases disease risk, but further loss of species decreases disease risk again. In this thesis, I have quantified host-feeding relationships of adult and immature Neotropical ticks, many of which (in the case of larvae and nymphs) were largely unknown. I have shown that adult ticks tend to be highly host-specific, particularly to larger-bodied vertebrates, whereas immature ticks appear to have broader host-use patterns. I found that ticks are susceptible to local host-tick coextirpation, and that the relationship between biodiversity loss and tick-borne disease risk is non-linear. My results emphasize the importance of directly assessing host community composition and suggest that the presence of specific (reproduction) hosts are a more important factor than species richness per se for tick population and tick-borne disease dynamics.