Human urges to access land, land resources, and other humans drive the propagation of roads. Human access almost inevitably causes ecological disturbance. In order to understand disturbance trends, therefore, we need to understand the spatial evolution of human access; this spatial approach is more accessible and data rich than are approaches rooted in behavioral science. There is a substantial geographic literature on road network evolution, developed primarily in relation to development of connections between population centers and the outward growth of urban areas. What is outward to a city is inward to watersheds and habitats; roads are invaders, facilitating incursion by the most invasive of species—Homo sapiens. Once a road is in place there is little impediment to the further import and application of all manner of mechanization. In this sense, roads beget roads and invasion begets further invasion. Given the ubiquity of this process it is remarkable that metrics to measure invasion have only recently been introduced. Physical presence of roads introduces one set of disturbances; traffic adds others. In backcountry settings and many ecological analyses, traffic information is not obtained in spite of broad recognition of its importance and in spite of ease of instrumental data acquisition. More surprisingly, road maps sufficient to support spatial analysis are often not available. Network structure determines traffic distribution and ease of access (generally measured in time) on roads themselves; terrain and vegetation determine accessibility of land between roads. In combination, these determine overall human accessibility. Weather and climate are readily introduced into accessibility analyses. We illustrate computational methods and results for all these models.
We are now in a position to look for ecological responses to trends in accessibility and, therefore, to assess the degree to which inaccessibility conveys protection from road induced ecological effects. Timescales of road network change and climate change are comparable, and introduction of road network evolution into long term ecological models, such as those of carbon sequestration, would provide estimates of the balance of change driven indirectly by climate and directly by local human action.