Increasing the function in distance-based functional connectivity assessments: a modified spatial interaction model (SIM) approach

Assessments of connectivity are subject to the limitations imposed by the technique used to model species movement. Least-cost path analysis is one such commonly applied technique that has a number of associated limitations that are often discounted, such as the assumption of individual omniscience or symmetrical movement between patches. We contend that not accounting for methodological limitations may lead to inaccurate assessments of connectivity, and thus there is a need for more robust and adaptable approaches. Using simulated data and in a case study on Ord's kangaroo rat (Dipodomys ordii), we present an approach that uses effective distance, in this case calculated from least-cost paths, in combination with a spatial interaction model (SIM), which allows one to incorporate additional landscape characteristics and interactions influencing movement, thereby overcoming key assumptions and limitations associated with the least-cost framework. We show how various factors influencing connectivity can be incorporated, how outputs from the SIM can be used to quantify connectivity, how outputs from different models may be compared, and, importantly, that in both a simulated and empirical case study application, the assessment of functional connectivity is sensitive to small changes to the model.