The impacts of host community composition on Lyme disease risk in Scotland

Gandy, Sara Louise (2020) The impacts of host community composition on Lyme disease risk in Scotland. PhD thesis, University of Glasgow.

Abstract

Emergence of zoonoses are driven by multiple factors, ranging from climate change to urbanization and human behaviours. Because many zoonotic pathogens are maintained in wild reservoir hosts, these factors of emergence may affect disease risk by changing host community parameters. Thus, it is important to understand the effect of host community composition on disease risk. This is particularly relevant for vector-borne zoonoses as host community composition might affect both reservoir host and vector populations. In the northern hemisphere, Lyme disease, a zoonosis caused by the bacterial complex Borrelia burgdorferi sensu lato, is the most prevalent vector-borne disease affecting humans. Transmitted by Ixodid ticks, its epidemiological cycle is complex and depends on environmental factors and host community composition, which together, influence both tick survival and the prevalence of B. burgdorferi s.l. Small mammals are competent reservoir hosts for B. afzelii, a genospecies belonging to the B. burgdorferi s.l. complex, while birds can transmit B. valaisiana and B. garinii, two other genospecies in the complex. Thus, pathogen prevalence will depend on the abundance of competent reservoir hosts in an environment. Deer species, on the other hand, are non-competent hosts and could, theoretically, lower B. burgdorferi s.l. prevalence by diverting immature ticks from feeding on competent reservoir hosts, a concept called the dilution effect. However, deer also act as the main tick reproduction hosts, feeding adult female ticks, and can therefore maintain high tick populations. In this thesis, I used a range of approaches (large-scale cross-sectional survey, deer exclosure and wood mouse supplementary feeding experiment) to test hypotheses and understand how host community composition drives Lyme disease hazard (defined here as the density of infected nymphal ticks) in Scotland and to investigate the drivers of host communities.

I investigated how the ratio of small mammals to deer affected Lyme disease hazard by conducting a large cross-sectional survey across sites selected to specifically cover the full range of deer densities (Chapter 2) and collecting data on host abundance (deer and small mammals), tick density and pathogen prevalence. There was a positive association between the density of infected nymphs (DIN) for B. afzelii and deer density, regardless of small mammal abundance. I also observed a positive association between deer density and quest in deer density was also negatively associated with human activity, which therefore has the potential to impact Lyme disease hazard.

Having shown how the full range of deer densities shapes Lyme disease hazard, which also suggested that deer may affect small mammal densities (Chapter 2), I then wanted to investigate further the possibility that high deer densities may affect Lyme disease by causing ecological cascades, through an impact on vegetation and small mammals (Chapter 3). I used an experimental design consisting of replicated fenced deer exclosures to investigate the effects of high deer density versus deer absence on Lyme disease hazard through ecological cascades. Consistent with my predictions, high deer density plots had 18 times more questing nymphs compared to plots where deer were absent. High deer density plots also had 13 times fewer small mammals and were associated with shorter and sparser vegetation and shorter trees, highlighting the impacts of browsing pressure by deer on small mammal communities. I found that the reduction in competent host abundance had repercussions on B. burgdorferi s.l. prevalence in questing ticks, which was twice as high in deer exclusion plots (2.2% in deer exclusion plots vs 1.0% in high deer density plots). Despite the negative effect of high deer density on B. burgdorferi s.l. prevalence, DIN was five times higher in high deer density plots. These results demonstrate that the positive effect of deer on tick density can outweigh their negative effect on B. burgdorferi s.l. prevalence caused by a dilution effect and through their negative effects on small mammals.

In chapters 2 and 3, I found that deer appeared to be more important than small mammals in driving DIN, through their strong role as tick reproduction hosts, irrespective of any dilution effects on prevalence. However, few of my sites had high small mammal densities while many had high deer densities. I therefore wanted to test the effect of a variation in small mammal abundance on ticks, B. burgdorferi s.l. prevalence and DIN using a supplementary feeding experiment to increase small mammal densities. More specifically, the objective for this study (Chapter 4) was to understand the effects of a change in food resources on small mammal populations and the repercussions on Lyme disease hazard. Fluctuations in resources availability are common occurrences in nature so it is important to understand the consequences this could have for disease hazard. This involved an experimental design where two trapping grids out of four were supplemented with food for two consecutive years. Density of live-trapped wood mice (Apodemus sylvaticus) was twice as high in food supplemented grids during the year of treatment. Food supplementation per se, which reflects what would happen after a mast seeding event, did not affect the density of nymphs, B. burgdorferi s.l. prevalence or DIN. However, there was a positive correlation between wood mouse abundance and questing nymph abundance the following spring. I hypothesised a negative association between wood mouse abundance and the prevalence of the bird associated B. garinii and B. valaisiana (dilution effect where more ticks would be feeding on mice instead of birds) but I did not find evidence for that. Unexpectedly, there was a positive association between wood mouse abundance and DIN for B. garinii and B.valaisiana the following spring. This could occur either if mice acted like deer did in Chapters 2 and 3, as a tick amplification host, since they increased nymph density but not B. garinii and B. valaisiana prevalence and/or if the food supplementation that increased mice also increased birds, which I did not measure. Even though food supplementation per se did not have a direct impact on DIN, wood mouse abundance had a positive effect on tick density and DIN for B. garinii and B. valaisiana and these results highlight how variations in food resources can affect small mammals and Lyme disease hazard.

The cross-sectional study (Chapter 2) is the first, to my knowledge, to test how the ratio of competent reservoirs hosts (small mammals) to non-competent hosts (deer) affects Lyme disease hazard, whilst also specifically selecting sites with the full range of deer densities to robustly test the dilution effect. While I predicted that DIN should be modulated by variations in the density of both hosts, my results really pointed to deer density being the key component driving Lyme disease hazard. This result, which was further confirmed in Chapter 3, could be used for deer management decisions in Scottish woodlands. This thesis also demonstrates the complexity of Lyme disease ecology: not only due to the multiple host types and multiple pathogens in the burgdorferi s.l. complex, but also in terms of how the hosts themselves interact and affect each other’s distributions and densities. There is now clearly a need for further research to better understand the mechanisms driving Lyme disease hazard, including interspecific interactions and other drivers suggested in this thesis, such as human disturbance and food resources, that affect host abundance. Combining experimental designs and large-scale surveys allowed a better understanding of the effects of deer and small mammals simultaneously on Lyme disease. Indeed, the experimental designs (Chapters 3 and 4) offered the chance to vary the density of one host species and observe the cascading effects on other hosts, tick density and burgdorferi s.l. prevalence. Results from the cross sectional-survey (Chapter 2) not only corroborated the key associations observed but, importantly, allowed me to robustly test the dilution effect by examining the effects of a wide range of deer densities in the natural environment on DIN. The results obtained can be used to understand and predict how environmental changes (e.g. increase of resources, increase of human activity) might impact host species and Lyme disease hazard. Furthermore, they could be used to predict the effects that a change in deer management might have on tick density and DIN.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Keywords: Lyme disease, Ixodes ricinus, dilution effect, deer, small mammals, reproduction host, density of infected nymphs
• Subjects: Q Science > QH Natural history > QH301 Biology; Q Science > QL Zoology; Q Science > QR Microbiology > QR180 Immunology
• Colleges/Schools: College of Medical Veterinary and Life Sciences > Institute of Biodiversity Animal Health and Comparative Medicine > Transmission Dynamics
• Supervisor's Name: Biek, Dr Roman and Gilbert, Dr Lucy
• Date of Award: 2020
• Embargo Date: 30 April 2022
• Depositing User: Dr Sara Gandy
• Unique ID: glathesis:2020-81708
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 12 Oct 2020 10:35
• Last Modified: 12 Oct 2020 10:35
• URI: https://theses.gla.ac.uk/id/eprint/81708