epizootic

epizootic

An R package for process-explicit models of wildlife disease

July Pilowsky

What is machine learning?

What is a process-explicit model?

What do you need to make one?

• Theory about how the study system works
• This is all you really need, because process-explicit models can be data-free

OK, July, but not all models can or should be data-free

• Data on processes (such as nitrogen fixation rate or replication fidelity)
• Data on drivers (such as temperature or % impervious surface)
• Observed patterns that can be compared to model outputs for validation

How do these models apply to disease ecology?

Introducing epizootic

The structure of epizootic

Case study: house finch conjunctivitis

epizootic in practice

epizootic in practice

Summer 1940	Winter 1993

Processes and drivers considered for house finch conjunctivitis

• Land use change (urbanization)
• Climate change
• Density-dependent dispersal
• Local extirpation by the disease

What questions could we answer using epizootic?

• Does urbanization facilitate the spread of wildlife disease?
• Does topography/elevation slow the spread of wildlife disease?
• Does hunting change the dynamics of wildlife disease?
• Suggest some I haven’t thought of yet!

Roadmap for epizootic

• Complete the house finch conjunctivitis case study
• Create a tutorial so it is easier for other people to use
• Add functionality for dispersal that differs by season and for the inclusion of hunting (?)
• Case study in vector-borne system (?)