Overview:
The industrialization of farming empowers farmers to keep pace with the ever-increasing demands of consumers. To most people, this industrialization is not associated with the rapid spread and evolution of infectious diseases. However, domestic livestock populations differ from unmanaged animals in fundamental ways that are highly conducive to pathogen evolution. Industrialization has been associated with increases in livestock population size and density and a decrease in rearing periods. Today, livestock accounts for about 20 per cent of the total terrestrial animal biomass. and while populations in the wild fluctuate in response to available resources, pathogens and predators, whereas domestic populations are managed to remain stable. As a result, pathogens are provided with globally abundant but genetically similar, densely packed hosts, promoting rapid disease spread. Concurrently, the management of consistent livestock populations inhibits the fitness costs a pathogen would usually pay for increased host mortality, therefore promoting higher pathogen virulence.
In much of my research I explore the relationship between modern agricultural practices and the evolution of infectious diseases affecting a range of domestic and agricultural animal populations. I study this relationship across a range of pathogens and agricultural hosts, and broadly aim to answer the question: how do modern agricultural management practices impact infectious disease burden, the risk of disease outbreaks as well as the evolution of pathogen virulence?
I am also interested in general how spacial structure and host availability influences disease ecology.
Marek’s disease virus and industrialized poultry farming
Collaborators: Troy Day, Scott Greenhalgh
Poultry farming has undergone significant changes since the 1950s, when industry wide intensification began. Today, broiler chickens (chickens raised for meat) are raised with tens of thousands of genetically similar birds in crowded barns, with a lifespan as short as six weeks. These changes to animal husbandry have been implicated in the continual increase in virulence of Marek’s disease, a viral disease of poultry. Marek’s disease is highly contagious and economically important, costing an estimated annual $1–2 billion to the poultry industry globally. There is no treatment for Marek’s disease and mortality rates for unvaccinated birds can be as high as 100%. Over the past 60 years, the Marek’s disease virus has experienced continual evolution towards greater virulence. Each mutation towards greater virulence has been accompanied by a greater ability to overcome the Marek’s disease vaccine’s induced immunity. This has resulted in a succession of new vaccines offering only temporary protection between iterations of mutations, making it exceptionally difficult to manage.
Figure 1: Broiler farm model schematic.
In our work, we investigated how poultry husbandry practices contribute to the evolution of Marek’s disease towards greater virulence. We developed a mathematical model that tracks the spread of Marek’s disease on a single industrial broiler farm (Fig.1). The model captures both the within-cohort dynamics (rearing of the chickens) as well as the inter-cohort dynamics (cleaning and restocking of the barn) with the use of impulsive differential equations. Our findings indicate that the duration of time chickens spend on the farm has important evolutionary consequences. The shorter the cohort duration, the more evolution favors virulent strains of the virus. Additionally, we found that intensive cleaning between cohorts of consecutive flocks has a similar affect – more cleaning means selection for strains of higher virulence. Interestingly and unexpectedly, we found that increasing the stocking density on a farm, while promoting disease spread, selects for less virulent strains of the virus. This finding highlights a management dilemma, as efforts to minimize disease spread and burden also promote the evolution towards greater virulence.
In Canada, a shift from conventional laying hen cages, known as battery cages, to alternative enriched cages or free-range systems is underway in response to consumer demand. We have been able to directly inform policy makers with our mathematical model of Marek’s disease. I am working with the Egg Farmers of Canada to evaluate the economic impact of Marek’s disease on egg production through the use of alternative housing systems.
Controlling bovine tuberculosis when there is a wildlife reservoir.
Collaborators: Matthew Silk, Mike Boots, Darren P. Croft, Richard J. Delahay, Dave Hodgson, Robbie A. McDonald, Nicola Weber
Bovine tuberculosis, bTB, while rare in most developed nations, continues to be a major animal and human health issue in the UK. Control of bTB in the UK is made difficult due to a wildlife reservoir of European badgers (Fig 3b). Culling of badger populations has been used to control the spread of bTB to cattle, but has proved unpopular and controversial with the public. Moreover, the scientific merits of the practice remain unclear. Whilst intensive badger culling has been shown to reduce the incidence of bTB in cattle in culled areas, it has also been associated with an increase in incidence in cattle in adjacent areas. The increase is incidence is postulated to be due to cull-induced perturbations to the badger populations. In much of the UK, badgers live at higher densities than in the rest of the species’ range; they live in territorial social groups that share communal dens and interact infrequently with individuals outside of their den. Quantifying how the unique social organization of the badger populations alters disease dynamics in this system is crucial in the development of successful strategies for controlling the spread of bTB in badgers and onwards transmission to cattle.
Figure 3: a) Badger social contact network, nodes colored by degree, b) European badger (photo by Keith Silk).
We address this question through the analysis of an exceptionally detailed social network dataset for a high-density population of European badgers that are naturally infected with bTB (Fig 3a). We used disease simulations to examine the implications of social group structures for both the risk and size of epidemics in networks generated directly from the empirically-derived European badger contact network. We show that the unique high-density subgrouping of the badger populations results in lower disease risk, whereby outbreaks are no less likely, but are smaller in size. This is significant in that the ongoing badger cull threatens to displace badgers from their dens, changing the population social structure which can consequently increasing the number of badgers infected with the disease. Our work suggests that the controversial badger cull could in fact be inhibiting management.
The disease consequences of honeybee apiculture intensification.
Collaborators: Mike Boots, Lewis Bartlette, Lena Wilfert, Keith Delaplane, Berry Brosi and Jaap De Roode
There has been widespread alarm and public concern regarding the decline in honeybee populations globally due to Colony Collapse Disorder, a phenomenon in which the majority of worker bees disappear from a colony. The alarm is not unfounded, as it is estimated that one in every three bites of food eaten globally depends on pollinators, mainly honeybees. Honeybee health and the apicultural industry is under threat from a variety of pressures including disease burdens caused by parasites and pathogens. As a result, there has been growing uncertainty around the short-term economics of pollinator-dependent agriculture and long term global food security. A growing body of literature is documenting the damage emerging or re-emerging diseases are causing to apiculture. However, a key outstanding question is how modern agricultural intensification and novel agricultural practices impact the emergence and epidemiology of infectious disease in bees.
Figure 2: a) Single bee hive, b)-d) bee hive configurations for a 9-hive apiary. Nodes represent hives and edges represent possible routes for between hive disease transmission.
We are the first to build a multi-hive mathematical model, on the scale of a single apiary (see Figure 2b-d for model configurations). We use our model to examine how apicultural intensification impacts honeybee pathogen epidemiology. We specifically examined the epidemiological consequences of increasing the number of hives within an apiary and changes to hive configurations. Our results present a perhaps counterintuitively positive picture for apiculture intensification and its consequences on disease burden. Changes in size, structure, and drift – reflecting some of the critical aspects of intensification in this system – have little effect on the severity of disease across an apiary. This is due to the unique ecology of honeybees and the rapid spread of disease within hives and can not necessarily be generalized to other livestock systems.
Perception Kernel and Vector Movements
Collaborators: Janis Antonovics
… coming soon 
