An Assessment of an Agri-environmental Policy on Landscape Connectivity for Wildlife

Abstract

Agricultural land conversion has resulted in the loss of 95-98% of tallgrass prairie habitat in the US. The rapid expansion and intensification of agricultural practices has played a pivotal role in meeting the ever-growing global demand for food. However, this intensification has come at an undeniable cost to the environment, especially in the form of habitat loss. In a high intensity agricultural system such as the US corn belt, agri-environmental policies can have a significant impact on remnant ecosystems. One common policy in this regard is the implementation of riparian buffer strips, designed to reduce pollution and soil runoff from agricultural fields into adjacent waterways. However, these strips also have the potential to provide critical habitat for wildlife, fostering landscape connectivity and aiding in the conservation of biodiversity. While the federal government provides incentives to encourage the creation and maintenance of buffer strips, the only state in the corn belt to require them is Minnesota. I built a machine learning model to classify high-resolution aerial imagery of this region over a time series from 2008, before policy implementation, to 2021, after implementation. The model separated agricultural and developed lands from permanent vegetation with the assumption that permanent vegetation will support the dispersal of wildlife more effectively than corn and soy monocrops. These classified maps were quantified using landscape connectivity metrics and compared between MN and other corn belt states, which collectively served as a synthetic control. The synthetic control model estimated high, positive impacts of buffer strip policy on landscape connectivity metrics.