Statistical models for incidence of Coccidiosis parasites in mink

In this report we present statistical models that describe the incidence of Coccidiosis parasites in mink in thirty mink farms in Denmark during the period April to October 2016. The dataset was obtained by asking the farmer at each mink farm to select 5 sows and collect a fecal sample from the sows on 16 dates during the period April to October. Once the sows give birth to kits (usually at the end of april), a pooled fecal sample from the kits belonging to each sow was also collected on the same dates as the sows. All the fecal samples were then analysed for presence of two types of Coccidiosis-causing parasites: Eimeria and Isospora. If Eimeria was present, the sample was examined for the presence of three different types of Eimeria, called tyk, tynd and lille. In addition, the fecal sample was classified by the farmer according to a feces score (hereafter F score) ranging from 1 to 5.
Various characteristics of the mink and the farms were also included in the dataset, including from which of three feed suppliers the farm obtains the feed, birth date of the kits, litter size, age of the sow. The aim of this project was to use statistical models to investigate which factors can account for the incidence of Eimeria (and its three subtypes), and Isospora, as well as the F score, observed during the study. This means there were 6 response (independent) variables to investigate: Eimeria, tynd, tyk, lille, Isospora and F score.