Challenges in characterizing a parcelized forest landscape: Why metric, scale, threshold, and definitions matter

Abstract Several metrics have been cited in the literature as being useful characterizations of forest land parcelization. Yet no agreed-upon standard measure exists which creates difficulties in identifying where parcelization is occurring as well as comparing the magnitude of its occurrence across different studies and geographic regions. We evaluated three existing (average parcel size, Gini coefficient, Shannon Entropy index) and a new metric (adjusted mean parcel size) for their usefulness in characterizing the extent to which a private forested landscape has become parcelized. We applied these measures to 410 forested townships in a contiguous, six-county area of northern Minnesota encompassing nearly 3.64 million hectares. Our analyses show that each metric typically describes a different pattern of parcelization and highlight problems owing to the fact that each metric appears to capture different aspects of ownership patterns within a landscape. We demonstrate the choice of metric, landscape scale, spatial and physical ownership features, and threshold for determining when a landscape is parcelized can greatly influence our conclusions regarding parcelization. Thus, careful consideration must be given to these factors when attempting to analyze a parcelized landscape, and caution is urged in interpreting and comparing parcelization studies where one or more of these factors vary.