Effective management action for biodiversity in farmland can often be complex given the multitude of species' needs and various habitat requirements. Biodiversity and land cover data from a previous study in eastern Ontario (Fahrig et al., 2015) was used to conduct random forest regression as a decision tool. This analysis was conducted for a variety of landscape metrics and their effect on alpha, beta, and gamma diversity and abundance of birds, bees, butterflies, syrphids, spiders, carabids, and plants. Main findings of the research supported the initial assertion by Fahrig et al. (2015) to make field sizes smaller in the interest of biodiversity conservation. It was also found that field types should be variably distributed, which has implications for biodiversity and land use management in eastern Ontario.