Characterizing forest fragmentation: Distinguishing change in composition from configuration

Forest fragmentation can generally be considered as two components: 1) compositional change representing forest loss, and 2) configurational change or change in the arrangement of forest land cover. Forest loss and configurational change occur simultaneously, resulting in difficulties isolating the impacts of each component. Measures of forest fragmentation typically consider forest loss and configurational change together. The ecological responses to forest loss and configurational change are different, thus motivating the creation of measures capable of isolating these separate components. In this research, we develop and demonstrate a measure, the proportion of landscape displacement from configuration (P), to quantify the relative contributions of forest loss and configurational change to forest fragmentation. Landscapes with statistically significant forest loss or configurational change are identified using neutral landscape simulations to generate underlying distributions for P. The new measure, P, is applied to a forest landscape where substantial forest loss has occurred from mountain pine beetle mitigation and salvage harvesting. The percent of forest cover and six LPIs (edge density, number of forest patches, area of largest forest patch, mean perimeter area ratio, corrected mean perimeter area ratio, and aggregation index) are used to quantify forest fragmentation and change. In our study area, significant forest loss occurs more frequently than significant configurational change. The P method we demonstrate is effective at identifying landscapes undergoing significant forest loss, significant configurational change, or experiencing a combination of both loss and configurational change.