Authors

F Occhibove¹; K Kenobi¹; C Risley¹;  ¹ Aberystwyth University, UK

Discussion

Tick-borne diseases include several of high importance both for human health and for conservation; furthermore, ticks can bridge host populations and these diseases have therefore been a matter of huge interest in the context of the nexus between wildlife disease ecology and public health.

In particular, rodent-tick disease systems, due to the rodent biology/ecology and the common host-generalism of the ticks, have been extensively studied to identify the connection between biodiversity and disease transmission. There has been widespread debate on the relative importance of the dilution versus amplification effects; i.e. does higher biodiversity decrease or increase disease transmission?

This eco-epidemiological modelling study aimed to investigate dilution/amplification in two different host-parasite-pathogen systems by varying community composition in order to highlight its effects on model outputs such as vector numbers and disease prevalence in the presence of parameter variation.

The model employed consisted of a deterministic single-vector multi-host compartmental model, also including ecological relationships with non-host species such as competition and predation, in order to describe disease transmission in a hypothetical Welsh woodland rodent community. The dilution effect was investigated by assembling a progressively more complex community. The two systems chosen for the analysis were Ixodes ricinus-Lyme disease and I. trianguliceps-Babesiosis.

Sensitivity analysis was performed on parameters that were hypothesised to affect model outputs but could not be directly estimated (e.g. efficiency in transmission/competence, host-vector encounter rate, feeding/moulting probability of success).

To improve model predictions, where relevant, parameters were estimated from field data. Rodent live-trapping was conducted in Welsh woodlands across two seasons for three consecutive years.

Modelling results confirmed that community composition and inter-specific dynamics strongly affect disease transmission; hence, in similar studies, the inclusion of ecological relationships within models would be expected to obtain more realistic predictions. The parameters most affecting the systems were among those where estimation in the field was more difficult, demonstrating that more empirical research is needed to reduce parameter uncertainty. Finally, dilution or amplification effects might not be mutually exclusive and their detection depends on the result under consideration, which has to be selected according to the aim of the eco-epidemiological study (e.g. public health or wildlife management/conservation). In the presented Lyme disease case, for example, more complex communities have fewer infected rodents (i.e. evidence for the dilution effect); however, increasing the number of species in the community led to an increase in the number of infectious nymphs, the metric mostly used as a proxy of human disease risk, amplifying it.