Skip to main content
Dryad logo

No net effect of host density on tick-borne disease hazard due to opposing roles of vector amplification and pathogen dilution

Citation

Gandy, Sara et al. (2022), No net effect of host density on tick-borne disease hazard due to opposing roles of vector amplification and pathogen dilution, Dryad, Dataset, https://doi.org/10.5061/dryad.rv15dv49r

Abstract

To better understand vector-borne disease dynamics, knowledge of the ecological interactions between animal hosts, vectors and pathogens is needed. The effects of hosts on disease hazard depends on their role in driving vector abundance and their ability to transmit pathogens. Theoretically, a host that cannot transmit a pathogen could dilute pathogen prevalence but increase disease hazard if it increases vector population size. In the case of Lyme disease, caused by Borrelia burgdorferi s.l. and vectored by Ixodid ticks, deer may have dual opposing effects on vectors and pathogen: deer drive tick population densities but do not transmit B. burgdorferi s.l. and could thus decrease or increase disease hazard. We aimed to test for the role of deer in shaping Lyme disease hazard by using a wide range of deer densities while taking transmission host abundance into account. We predicted that deer increase nymphal tick abundance while reducing pathogen prevalence. The resulting impact of deer on disease hazard will depend on the relative strengths of these opposing effects. We conducted a cross-sectional survey across 24 woodlands in Scotland between 2017 and 2019, estimating host (deer, rodents) abundance, questing Ixodes ricinus nymph density and B. burgdorferi s.l. prevalence at each site. As predicted, deer density was positively associated with nymph density and negatively with nymphal infection prevalence. Overall, these two opposite effects cancelled each other out: Lyme disease hazard did not vary with increasing deer density. This demonstrates that, across a wide range of deer and rodent densities, the role of deer in amplifying tick densities cancels their effect of reducing pathogen prevalence. We demonstrate how non-competent host density has little effect on disease hazard even though they reduce pathogen prevalence, because of their role in increasing vector populations. These results have implications for informing disease mitigation strategies, especially through host management.

Methods

We used 24 sites located in Aberdeenshire, Northeast Scotland. An index for deer density was estimated the first year of data collection using the standing crop plot count method (Mayle et al., 1999). Rodent abundance was estimated at each site during the first year of data collection (2017 for 15 sites and 2018 for nine sites) using two methods. First, rodent abundance was quantified at each site by live-trapping using non-selective Sherman traps. The second method used to assess rodent abundance was by recording vole signs (tunnels and holes) at each site. We then created a simple index of rodent abundance which combined data from both methods. First, we classified each site as having either low or high rodent abundance from each method. The data from each method exhibited a bimodal distribution with a clear gap between low and high density . If a site scored a high category for at least one index, it was defined as a high rodent abundance site whereas sites which scored low categories for both indexes were identified as low rodent abundance. 

Questing nymphs were collected three times a year (May, July and September) for each year following host density estimation (2018 and 2019) using a standard blanket dragging method . A white 1 m x 1 m square of fleece blanket material was dragged over vegetation along 10 m long transects. At each site, twenty transects were randomly surveyed and separated from each other by at least 20 m. Nymph ticks on the blanket were counted, collected and kept at -20°C for pathogen analysis. Woodland type (coniferous, deciduous or mixed woodland) was recorded at the site level while ground vegetation was classified into four categories: (1) grasses and herbaceous species, (2) Ericaceous species (Calluna and Vaccinium), (3) moss species and (4) bracken and ferns. The dominant vegetation category (that with the most cover) over each transect was recorded. Vegetation height was measured at the beginning (1 m), middle (5 m) and end (10 m) of each transect, and the mean for each transect was included as a continuous variable in analysis of tick density as it can affect dragging efficiency. Tick surveys were conducted between 0900 hours and 1800 hours and air temperature, relative humidity and time were also recorded for each transect, as these may affect the proportion of nymphs questing. For analysis we also used the rainfall recorded the day before ticks were collected from the nearest weather station at Aberdeen Airport (http://rp5.co.uk/Weather_archive_in_Aberdeen_(airport),_UK).

Nymphs were extracted individually using an ammonia extraction method. Borrelia burgdorferi s.l. was detected from samples using a qPCR protocol on fragments of OspA genes based on the protocol described by Heylen et al., 2013.

All statistical analyses were performed in the software R version 3.5.1 (R core team, 2013) using the glmmTMB, lme4 and MuMIn packages. 

Usage Notes

The data files can be open using Microsoft Excel.