Modelling systems for managing bee disease: the epidemiology of European Foulbrood
Lead Research Organisation:
Newcastle University
Department Name: Sch of Biology
Abstract
This project will provide a step-change in our understanding of managed pollinator disease. We will use a combined modelling and molecular approach to investigate the dynamics of European Foul Brood (EFB) as an exemplar of endemic brood disease of honey bee colonies using historic data derived from long-term monitoring of apiaries in England and Wales. We will utilise a program of statistical, analytical and spatially explicit modelling to address the problem. Statistical modelling approaches will be used to identify putative covariates involved in the epidemiology of disease (e.g. land use, weather, management practices) (Newcastle); analytical modelling approaches will be used to investigate the role of transmission processes in determining the epidemiology of disease (Warwick & Bath); and spatially explicit models to investigate spatial spread of disease in the context of investigating the efficacy of different practical control measures (Warwick & Newcastle). The modelling will be parameterised using historic datasets which include the timing and reported incidence of EFB distribution in honey bee apiaries across England and Wales (Fera). Molecular approaches based on microsatellite markers and comparative genomics will be employed to characterise host and parasite diversity (Fera & Bath) for use as additional covariates in the statistical, analytical and spatially explicit models exploring the epidemiology of EFB in relation to host resistance. These data will be used for the testing and validation of the theoretical and spatially explicit models. We (Fera & Bath) have, in collaboration with the Sanger centre in Cambridge, already generated a draft genome sequence for M. plutonius. These data will greatly facilitate the identification of suitable markers for the characterisation of large and representative population samples and will also shed light on the genes responsible for virulence, and how pathogenesis proceeds in the bee host. EFB will provide a paradigm which we can test against other pollinator diseases. For example, developed models will be used to investigate the epidemiology of 14 honey bee diseases collected across 5000 apiaries as part of an ongoing Defra funded monitoring programme (Fera). Dissemination of project results is explicit within the project framework and includes, the production of a list of key end-users, stakeholder workshops, bi-annual project newsletters, reporting in industry literature, a disease management summary document and conference attendance. The modelling analytical and spatially explicit models developed within this project will act as tools to guide strategy in the face of a plethora of disease threats for managed and wild pollinators.
Technical Summary
We will use a combined modelling and molecular approach to investigate the dynamics of the honey bee brood disease, European Foulbrood (EFB), as an exemplar of pollinator disease using historic data derived from long-term monitoring of apiaries in England and Wales. We will utilise a statistical, analytical and spatially explicit modelling program to investigate the epidemiology of EFB and to provide practical management solutions for this disease. Statistical modelling approaches will be used to identify putative covariates involved in the epidemiology of disease (e.g. land use, weather, management practices) analytical modelling approaches will be used to investigate the role of transmission processes in determining the epidemiology of disease; and spatially explicit models to investigate spatial spread of disease in the context of investigating the efficacy of different practical control measures. The modelling will be parameterised using historic datasets which include the timing and reported incidence of EFB distribution in honey bee apiaries across England and Wales. Molecular approaches based on microsatellite markers and comparative genomics will be employed to characterise host and parasite diversity for use as additional covariates in the statistical, analytical and spatially explicit models exploring the epidemiology of EFB in relation to host resistance. Molecular data from newly collected samples will be complemented by historical data generated from stored EFB test kits. These data will be used for the testing and validation of the theoretical and spatially explicit models. The models will be further tested on data on incidence of other pathogens present in 5000 apiaries across England and Wales (current Defra study) and a generic framework developed for investigating disease spread from these and exotic pathogens.
Planned Impact
This project will have tactical and strategic impacts at three levels. First, models will help to develop sustainable disease management practices that beekeepers can apply with guaranteed distribution to end-users via an existing framework of training and extension at the NBU. Second, this project will improve the regulatory risk based inspection programme for the control of bee pest and diseases across England and Wales by targeting approaches and procedures for disease control and driving future research. Third, this project will inform policy makers, including the Scottish Government, on the necessary strategies required to minimise the impact of endemic diseases and also ensure preparedness for emerging threats to bee health. Overall this project supports the aspirations of the Defra and Welsh Assembly Government's 10 year 'Healthy Bees' plan and will contribute significantly to our understanding of bee disease.
People |
ORCID iD |
| Steven Rushton (Principal Investigator) |
Publications
Budge GE
(2014)
Molecular epidemiology and population structure of the honey bee brood pathogen Melissococcus plutonius.
in The ISME journal
Mill AC
(2014)
Clustering, persistence and control of a pollinator brood disease: epidemiology of American foulbrood.
in Environmental microbiology
| Description | • K-function analysis indicated significant spatial clustering of EFB in all years, which for most years occurred at a distance of 150 km or less. (Objective O1a) • Kernel smoothing identified persistent hotspots of EFB infection in SE England between 1994 and 2001, after which time the focus shifted gradually westwards until 2008. (Objective O1a) • Cuzick-Edwards tests for clustering identified small clusters of 2 to 3 cases that did not span years. in 1/3 of all cases the closest apiary to an infected apiary was another infected apiary, but in only 3% of cases was the closest apiary to a non-infected apiary was infected. (Objective O1a) • Bayesian conditional autoregressive models were used to estimate the relative risk of cases of EFB in hexagons of 5km minimum diameter, with landcover, past infection, and weather data as explanatory variables. Areas of high relative risk were located mostly in the south of England. The number of Priority Inspection Programme inspections (PIPs) in the hexagon was the only consistently significant covariate even after accounting for spatio-temporal autocorrelation. A reasonable explanation for this result is that areas susceptible to high levels of disease are under a higher level of surveillance above and beyond that provoked by the presence of EFB cases. (Objective O1b) • Cox proportional hazard models were used to assess the extent to which the timing of infection events could be explained in terms of the covariates mentioned above. Random effects were incorporated into the models as frailties. Previous EFB in the hex and all three temperature variables were important drivers of the hazard of EFB. The southern region demonstrated an elevated hazard of infection, an apiary being anywhere between 8.2 and 17.2 times more likely to have an EFB infection in one of these counties than the average. (Objective O1c) • Structural Equation Modelling was used to test whether the variables in the hypothesised pathway of interaction were significantly interrelated by analysing their variances and co-variances. The number of EFB cases was found to be dependent on PIP; habitat and spring temperatures, which were themselves dependent on the altitude of the apiary. There were fewer PIPs at higher altitude, and the spring temperatures were lower. A trend towards semi-natural grassland from arable habitats was a positive predictor of EFB, as were low spring temperatures and low altitude. The more PIP inspections in the previous year, the fewer EFB cases, suggesting that PIP is a means by which infection by EFB can be controlled. (Objective O1d) • A cost-effectiveness analysis was performed on the historical data on EFB control, and predictive treatment scenarios were formulated based on converting all observed treatments to either Destruction, Shookswarm, or Antibiotic treatments. Destruction of infected colonies was predicted to be the only means by which EFB eradication can be achieved. However, destruction of colonies at an apiary is the most costly method of disease control for a beekeeper; and despite the predicted decrease in number of EFB cases the Net Present Value (NPV) of destruction was still the most expensive over sixteen years, even though costs were predicted to be lower than the current cost after ten years. (Objective O1e) • A spatially-explicit individual-based model was developed for the landscape-scale modelling of EFB. The model consisted of nearly 30,000 apiaries connected by physical proximity and common ownership. In a typical year the model simulated an average of 150 individual disease clusters, each cluster comprising 8.7 apiaries (range 2-65). The model was used to predict the impact of different management scenarios. Abolishing the current "amber zone" of enhanced surveillance and inspecting all apiaries within 10 km of a confirmed case broke up the annual cycles of infection and reinfection and resulted in an overall reduction of EFB over time. An alternate management scenario which doubled the number of overall inspections regardless of scheduling was shown to reduce the number of new infections but not lead to EFB eradication. (Objective O3a) |
| Exploitation Route | The research has guided management of EFB disease through assessing the current monitoring and treatment approaches used. |
| Sectors | Agriculture, Food and Drink,Environment |
| Description | Gain an understanding of epidemiology of European foulbrood, which has been disseminated to the beekeeper and apiculture industry through workshops and newsletters |
| Sector | Agriculture, Food and Drink |
| Impact Types | Economic |